Copyright (c) 1997 Tax Analysts
DECEMBER 22, 1997
DEPARTMENT: Special Reports (SPR)
CITE: 77 Tax Notes 1403
HEADLINE: 77 Tax Notes 1403 - SIMPLIFICATION: REPLACE THE PERSONAL EXEMPTIONS PHASE-OUT BUBBLE. (Section 151) (Doc 97-34073 (5 pages))
AUTHOR: Johnson, Calvin
University of Texas
CODE: Section 151
Professor Calvin Johnson calls for elimination of the section 151(d)(3) personal exemption phase-out bubble, to be replaced by an increase in the top two individual income tax rates.
Professor Calvin Johnson calls for elimination of the section 151(d)(3) personal exemption phase-out bubble, to be replaced by an increase in the top two individual income tax rates.
Calvin H. Johnson is a professor of law at the University of Texas. The author wishes to thank members of the ABA Tax Section, Tax Structure and Simplification Committee, and especially Professors Jim Maule and Reed Shuldiner, for helpful comments and discussion. The language, and especially the errors, are the author's sole responsibility and may not be attributed to the ABA Tax Section or its committees nor to anyone who volunteered to help or critique.
This is the first of a two-part series on simplifying tax by repealing and replacing the phaseout taxes of section 151(d)(3) and section 68. This part covers the 'personal exemption phase-out' of section 151(d)(3). The second part covers replacement of the 'itemized deduction limitation' of section 68.
Johnson notes that section 151(d)(3) takes away personal exemptions, as adjusted income increases over a range from $181,000- $304,000. The effect is a surtax on children and other dependents, under which the more children within the household, the higher the surtax. Marginal rates under the surtax, he finds, can approach 48 percent for taxpayers within the surtax bubble. The surtax disappears for highest-income taxpayers, however, so it is lower-income taxpayers who are asked to pay the higher marginal rates. The phaseout is also extraordinarily complicated, taking nine confusing lines of the individual tax return worksheets. This report advocates replacing section 151(d)(3) with an increase of approximately 1 percent in the top two brackets of section 1.
TABLE OF CONTENTS
Current Law. . . . . . . . . . . . . . . . . . . . . . . . . . . 1403
Reasons for Change. . . . . . . . . . . . . . . . . . . . . . . .1404
Explanation of Proposal. . . . . . . . . . . . . . . . . . . . . 1405
Revenue Neutrality. . . . . . . . . . . . . . . . . . . . . . . .1407
Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . .1407
1 As time goes by and new tax legislation is enacted, the Internal Revenue Code more resembles a bramble patch. One of the more complicated, least justified pieces of the bramble patch is the 'personal exemption phase-out' or 'bubble' surtax of section 151(d)(3) of the code. Section 151(d)(3) takes away the personal exemptions for the taxpayer and dependents, as adjusted income increases, over a range of income of moderately well-to-do taxpayers. The effect is a surtax, typically of 4 percent, on income within the range. The surtax disappears once personal exemptions have been fully lost. There is thus a bubble in the tax rates under which taxpayers with lower incomes face higher tax rates, and taxpayers with higher incomes get lower rates. The surtax is contingent, moreover, on dependent exemptions. The structure yields a 'family tax,' under which the more children or other dependents within the household, the higher the surtax. Smaller families get the lower tax rates and larger families get the higher. The exemption phaseout also adds nine incomprehensible lines to the individual tax return worksheets.
2 The law would be simplified, dramatically and elegantly, by replacing the personal exemption phaseout with an increase in the explicit, section 1 tax brackets, adding 1 percent to the top two tax brackets. The proposal would allow a drop in the real, combined highest tax rates. The surtax, typically of 4 percent of income, would disappear, replaced by the explicit 1 percent added tax. The proposal would end the structure that yields a 'family tax' on more dependents, it would end the 'bubble' structure under which it is the lower-income households that face the higher tax rates, and it would end the need for all nine worksheet lines. This report describes current law, then the reasons for change, and then the proposal itself, including the revenue estimate that sets the contours of the replacement proposal.
3 Section 151(d)(3) phases out personal exemptions by reducing the deductible portion of the exemptions by an increasing percentage of the exemptions over a range, e.g., for joint returns, of adjusted gross income, now between $181,000 and $304,000. /1/ Both the top and the bottom of the range are adjusted for inflation annually and the adjustments are announced before the year starts by a revenue procedure.
4 Within the range, section 151(d)(3)(B) has a silly notch or stairstep structure under which an extra dollar of income can yield an extra surtax that far exceeds confiscation of extra income the taxpayer has received. The taxpayer loses another 2 percent of all exemptions for self, spouse, and dependents for every $2,500 of added adjusted gross income or fraction thereof. The taxpayer will thus lose an entire added 2 percent of all exemptions of income, if he or she crosses a $2,500 line by even one dollar. The added tax will exceed the added income by many times, especially for income just above a new $2,500 line. /2/
5 If we average out the impact of section 151(d)(3), so as to ignore the stairstep effect, then the phaseout becomes like an extra tax or surtax of just below 1 percent per dependent or other personal exemption. Section 151(d)(3) takes away 2 percent of each exemption deduction for every extra $2,500 or part thereof within its range. Exemptions are now $2,650. /3/ The average impact is extra taxable income of 2 percent * $2,650 per $2,500 or 2% * 26.5/25, or 2.1% extra taxable income per exemption per dollar. The phaseouts operate across the 36 percent and 36.9 percent tax brackets. Within the 36 percent tax bracket, the impact is 2.1% * 36%, which equals a surtax of 0.76 percent per exemption. Within the 36.9 percent bracket, the impact is 2.1% * 39.6%, which equals a surtax of 0.83 percent per exemption. For a household consisting of a couple and three children (i.e., five exemptions), the surtax is between 3.8-4.2 percent of the taxpayer's income, again depending on the bracket. /4/ For household with 10 exemptions, the surtax could be 8.3 percent. The surtax disappears once all of the deductions for exemptions have been taken away.
6 When section 151(d)(3) is combined with the tax brackets found in section 1, the marginal rates will typically be 45 percent. For the three-child household, the exemption phaseout adds 4.2 percent to the 39.6 percent tax bracket for a top marginal tax rate of 43.8 percent. There is another phase-out tax, on certain itemized deductions, that will add another 1.18 percent to the marginal rate, yielding a marginal tax rate of 45 percent. /5/ Households with more than five exemptions will face even higher marginal tax rates. For the second highest tax bracket (36 percent), the two surtaxes for the three-child household yield a marginal tax rate of 41 percent. /6/ The highest, 45 percent marginal rate drops back down to 39.6 percent, however, once the two phaseouts have been completed.
7 The section 151(d)(3) surtax is triggered by adjusted gross income over the threshold, and not by taxable income. It thus ignores decreases to standard of living for reasons such as medical expenses, casualty losses, or investment interest. A taxpayer within the range of the phaseout who has significant medical expenses will save tax within his normal tax bracket for the deductible medical expenses, but will continue to pay the surtax as if the medical expenses were not there.
Reasons for Change
8 The adjustments in section 151(d)(3) are extraordinarily and unjustifiably complicated. The adjustments require nine lines in individual tax return worksheets, /7/ which are justifiably incomprehensible to the lay public. The phase-out taxes are complicated enough that the most expert tax economists in the country sometimes make mistakes describing them. /8/ The extraordinary complexity of section 151(d)(3) has prevented Congress from understanding or reviewing the provision. Nobody in Congress or on its staff has the time and patience to understand or to rationalize it.
9 The bubble structure under which the marginal tax rate on moderately high-income taxpayers is higher than that on the highest- income taxpayers is perverse. Lower-income households should not pay higher marginal tax rates than households with even more income. The highest tax rates should be reserved for the highest-income taxpayers. A phaseout of a benefit as income increases seems to have some intuitive appeal as if it were a reasonable remedy, but unfortunately, a bubble structure is an unavoidable aspect of a phaseout, no matter what benefits are involved. Phaseouts cause bubble taxes in impact whether it is welfare benefits, exemptions, or something else that is being phased out. The best that can happen under a phaseout is that the phase-out range is very wide so that the bubble tax will be low. If a phaseout can be avoided, it should be.
10 Greater tax under the section 151(d)(3) phaseout, moreover, is contingent both on more income and on more exemptions. Literally, section 151(d)(3) operates as a surtax on dependents' exemptions that is like a 'family tax' on children and other dependents. /9/ A household with only one taxpayer can face a bubble surtax of no more than 0.83 percent. A household with 10 dependents, however, can face a bubble surtax of 8.3 percent. No one in the debates or legislative history said that bigger families could bear more tax, all other things being equal, or that a tax on children and other dependents was intended or justified. When Congress expresses its value judgments directly, it seems to favor reducing tax for bigger families in recognition of the significant cost of raising children. /10/ It is difficult, in any event, to see how having more children or other dependents could mean that that household has less need and has greater ability to pay tax. The dependent exemptions seem to have been caught up in the provision only by accident. The real justification for the surtax is that they increase tax as income increases. That purpose would be accomplished more simply and without the side effects just by adjusting section 1 rates.
11 According to reports contemporaneous with enactment, section 151(d)(3) was adopted to allow politicians to raise revenue, while giving the appearance of not raising explicit tax rates. /11/ However, the objective impact of the personal exemption phaseout in section 151(d)(3) is to create extra tax for extra income, that is, to raise marginal tax rates for income within the range the phaseout operates. /12/ The maximum marginal rate including the phase-out taxes, i.e., 45 percent or above, is the tax rate that will affect the taxpayer's economic decisions within the range of the bubble and the 45 percent marginal tax is unnecessarily high.
12 Since the 45 percent (and higher) marginal tax rate from the combination of surtaxes and section 1 brackets is unnecessarily high, the phaseout produces unnecessary economic distortion. High marginal tax rates produce deadweight losses because taxpayers distort their behavior to avoid high marginal tax. Deadweight losses increase with the square of the increase in tax rate. /13/
13 Any transient political advantage from the indirection involved in section 151(d)(3) does not justify the unnecessarily high, real marginal tax rate, the impact of the tax as a penalty on families, or the excruciating complexity imposed on individual taxpayers indefinitely for years beyond enactment.
Explanation of Proposal
14 The current marginal tax rate structure would be considerably simplified and rationalized by repealing section 151(d)(3) and replacing the revenue with an explicit 1 percent increase in the tax brackets of section 1(a)-(d). The top bracket would be increased from 39.6 percent to 40.6 percent under the proposal. The second highest tax bracket would be increased from 36 percent to 37 percent. The proposal significantly reduces the highest combined marginal tax rates. The replacement of the phaseout by an explicit tax rate would reduce the highest tax rate from a combined rate of 45 percent or more to 40.6 percent. /14/ The replacement would reduce the second highest bracket from a combined rate of 40.86 percent or more to 37 percent. /15/ Dropping the highest marginal tax rates would improve economic efficiency. Replacing the revenue from section 151(d)(3) with explicit bracket adjustments would also end the need for all nine lines of the tax return worksheets now devoted to the exemption phaseout.
15 The highest, 40.6 percent tax bracket created by the proposal would not, however, drop for higher-income taxpayers, as the surtax does under the current bubble structure. The highest marginal rates would be reserved for taxpayers with the highest income. The extension of the 1 percent surtax indefinitely, beyond the current bubble, allows the proposal to reduce high marginal tax rates without shifting the burden of tax downward to poorer taxpayers relatively less able to pay. Poorer households are hurt more profoundly by any given dollar of tax. The revenue from extending the 1 percent increase indefinitely obviates the need for the 4 percent or more 'family tax' arising from the phaseout of dependent exemptions. Section 151(d)(3) currently reflects the unstated assumptions that a larger family should pay a higher marginal rate than a smaller family, for the same income, and that the less well-to-do household should pay a higher marginal rate than the richer household. Both assumptions, once expressed, need to be abandoned.
A. YIELD OF SECTION 151(d)(3). (,000 omitted)
bracket bracket sum; row Comments
1. Section average rate
151(d)(3) (notch effect
phase-out rate 2.10% 2.10% averaged out)
exemptions /17/ 2.95 2.76
income 6.20% 5.79% 1.*2.
Apportioning deduction reduction to brackets Row 4 through 7 /18/
tax year SOI 1994 Table
in bracket $ 90,733,103 3.5, col. (21)
5. National tax 4./$ 110,000
year per dollars (width of 36%
of bracket $ 825 bracket)
within bracket $ 82,300 $ 40,200
7. National tax
phaseout $ 67,897,500 $ 33,165,000 5.*6
8. Added tax
year $ 4,210,084 $ 1,921,659 7.*3
151(d)(3) yield $ 1,515,630 $ 760,977 $ 2,276,607 8.* bracket
B. YIELD OF REPLACEMENT: BRACKET ADJUST,, UP TO 37 PERCENT AND 40.6
10. National SOI 1994 Table
Tax by 3.5 col.
Bracket $ 32,663,917 $ 65,344,257 (22,25)
11. National SOI 1994 Table
Tax by 3.5 col.
Bracket $ 90,733,103 $ 165,010,750 $ 255,743,853 (21,24)
elimination ($ 4,210,084) ($ 1,921,659) ($ 6,131,743) =-8.
taxable income $ 86,523,019 $ 163,089,091 $ 249,612,11 11. -12.
rate increase +1%, +1%
15. Tax $ 865,230 $ 1,630,891 13.*14.
shift /19/ ($ 34,640)
tax $ 865,230 $ 1,596,251 $ 2,461,481 15.+16.
18. Change tax
in bracket ($ 650,400) $ 835,274 $ 184,874 10.-17.
16 A proposal justified solely as a simplification and improvement in tax structure needs to be revenue neutral. Simplification should not be used as a camouflage for significant shifts in the tax burden downward onto poorer taxpayers. The exact contours of any proposal to replace section 151(d)(3) necessarily depends on the accuracy of the revenue estimates. There are no publicly available estimates of the yield from section 151(d)(3), but reasonably reliable estimates can be constructed from published IRS statistics of income. Table 1 and the accompanying notes, explain the construction of the revenue estimates behind the replacement proposal. The logic of Table 1 is explained by short comments in the final column of each row and by longer footnotes where the column does not allow enough room.
17 Table 1 shows a small, unintended tax increase from the proposal -- $184,874,000. (See row 18, sum column.) The surplus or tax increase might be avoided, under the revenue estimates used here, by increasing the top tax rate by only 0.9 percent (instead of a full 1 percent) to only 40.51 percent. Alternatively, the small surplus might also be returned to current top bracket taxpayers by raising the dividing line between the top and second-highest brackets, extending the width of the second-highest bracket upward by about 0.5 percent, so that some now top-bracket taxpayers get to pay tax at the second lowest bracket. Once the proposal is made revenue neutral, returning the surplus to the top bracket, it will change tax paid as follows:
TABLE 1 (CONT'D). AFTER RETURN OF THE SURPLUS
Old 36% Old 39.6% sum change Comments
19. Change in
tax paid ($ 650,400) $ 650,400 ($ 0) 18. at revenue
change in tax -2.0% +1.0% 19/10.
18 In sum, Table 1 shows that the proposal will decrease tax paid in the old 36 percent bracket by 2 percent and will increase the tax paid in the top bracket by 1 percent. That is the result to be expected from replacing the bubble with a change in a top bracket that extends indefinitely.
19 The revenue estimates shown in Table 1 are a good faith effort to propose a revenue-neutral replacement for section 151(d)(3) on the basis of published revenue estimates, but on this issue, it is necessary to defer to the professional revenue estimators in Congress and Treasury, who have access to the latest data --- and quite possibly better techniques and fewer errors. The fine-tuning adjustments proposed here, whether in tax rates or bracket dividing lines, seem especially vulnerable to amendment in response to official revenue estimates.
20 The phase-out bubbles of section 151(d)(3) are extraordinarily complicated. Replacing them with an explicit, approximately 1 percent increase in the top two brackets would drop the real maximum marginal tax rates and end the perversity under which it is the lower-income taxpayers who must pay the highest marginal rates and the larger families that are deemed to be better bearers of tax than smaller families. The proposal is pro-family. It promotes efficiency by cutting unnecessarily high tax rates, without shifting tax to those less able to pay. It simplifies the tax law and yields a fairer rate structure. Who then could resist it?
/1/ Rev. Proc. 96-59, 1996-53 IRB 17, section 2.09 (2) (applicable to 1997 tax year).
/2/ With five exemptions at $2,650 and a section 1 rate of 39.6 percent, a dollar that crosses a $2,500 of line would cost an extra 5 * $2,650 * 2% * 39.6% tax, or $97.75 tax. Stated in percentage terms, that is a tax rate of 9,775 percent, far in excess of confiscation. There is, however, no further tax under the phaseout until another $2,500 line is crossed, so that the notch effect smoothes out into more reasonable tax rates as the taxpayers income increases through a $2,500 tier. The notch effect, while silly, probably does not have significant behavioral effects, because the extra tax is not material at the levels of income at which the phaseout operates and because taxpayers do not ordinarily have control over their income down to changes by $1 or so.
/3/ Rev. Proc. 96-59, 1996-53 IRB 17, section 3.09 (applicable to 1997 tax year).
/4/ For five exemptions the impact is 5 * 0.76% or 3.8 percent and 5 * 0.83% or 4.2%. Both the thresholds and the exemption amounts are adjusted for inflation.
/5/ Section 68 typically adds taxable income in the amount of 3 percent of adjusted gross income.
39.6% + 3% * 39.6% + 5 * 2.1% * 39.6% = 44.95%.
/6/ 36% + 3% * 36% + 5 * 2.1% * 36% = 40.86%.
/7/ See, e.g., Forms and Instructions 1040 (1996) at 22 (Instructions to line 36).
/8/ See., e.g., Henry Aaron & William Gale, Introduction, Economic Effects of Fundamental Tax Reform 1, 2 & 6 (1996) (describing section 68 and section 151(d)(3) as yielding 2 percent surtax, which is accurate only on the unusual assumption of only one personal exemption for a household).
/9/ Frank J. Murray and Major Garrett, ''Family Tax' Last Hitch in Budget Deal,' Washington Times at A1 (October 26, 1990).
/10/ See, e.g., Staff of Jt. Comm. on Taxation, General Explanation of the Tax Reform Act of 1986 at 15 (1987) (explaining increase in personal exemptions to recognize the significant costs of raising children).
/11/ See, e.g., Jeffrey H. Birnbaum & Alan S. Murray, Show Down at Gucci Gulch: Lawmakers, Lobbyists, and the Unlikely Triumph of Tax Reform, New York, New York Random House 89, 220 (1987) (phaseouts adopted, notwithstanding 'perverse result' that staff viewed as 'wacky' and ''gimmickry' inconsistent with the integrity of the Tax Code' to give appearance of lower tax rates).
/12/ See, e.g. Editorials. 'Rosty's Bubble,' Wall St. J. A18 Mar. 14, 1990 (condemning the bubble as disguised increases in marginal tax rate). Birnbaum and Murray (Wall St. Journal Reporters), supra note 11, at 89, identify the originator of the bubble idea as Richard Darman, director of the Reagan Office of Management and Budget. Attributing authorship to one's political enemies does indicate a healthy dislike of the bubble.
/13/ See, e.g., Joseph Stiglitz, Economics of the Public Sector 376 (1986) (magnitude of deadweight loss increases with the square of the tax increase).
/14/ With the 1.18 percent addition to the marginal tax rate caused by section 68, the proposal would reduce the highest marginal rate from 45 percent to 41.8 percent. See discussion accompanying infra note 20-25.
/15/ With the 1.08 percent addition to the marginal tax rate caused by section 68, the proposal would reduce the marginal tax rate from 41 to 38 percent.
/16/ Table 1 is based on data published in Internal Revenue Service, Statistics of Income Division, Individual Income Tax Returns 1994 (hereinafter 'SOI 1994'), which is the most recent, publicly available data. The Table 1 figures are 'static' revenue estimates that ignore taxpayer responses to the changes, but since the changes are projected to be revenue neutral and the proposals predominantly yield lower marginal rates without regard to industry or investment vehicle, it can be assumed that taxpayer responses will not be material, will increase revenue, or at least not decrease revenue.
/17/ SOI 1994, Table 1.4, col. (125) & (1). For the 36 percent bracket, the average exemptions per return were computed by dividing figures in col. (125) (number of exemptions) by col. (1) (number of returns) for the $100,000-$200,000 adjusted gross income row. For the 39.6 percent bracket, the row was adjusted gross income greater than $1 million.
/18/ In 1994, section 151(d)(3) began at $167,700 and ended at $290,200 for married couples. Rev. Proc. 93-49, 1993-2 C.B. 580, section 3.07. Section 151(d) means lesser deductions, but the impact depends on the bracket in which the deductions fall. Table 1 apportions the loss of deductions between the brackets on the assumption that income is spread evenly across the $167,700-290,200 range. (The assumption is inaccurate, but the same assumption is made in computing the replacement tax so the errors are offsetting. No more accurate assumption is available, moreover, from publicly available information.) National taxable income per dollar of bracket is calculated, so that national taxable income can be apportioned to each bracket according to the amount within the bracket that falls within section 151(d)(3). Total national taxable income (Row 4) is from col. 21 of Table 3.5 of SOI 1994 for the 36 percent bracket and the width of the 36 percent bracket in 1994 was ($250,000-140,000) or $110,000. (Rev. Proc. 93-49, 1993-2 C.B. 580, section 3.01.) National taxable income per dollar of bracket width is thus Row 4 divided by $110,00 or Row 5. Since the top of the 36 percent bracket was $250,000, that means that $250,000-$167,700 or $82,300 is within the 36 percent bracket (Row 6), which represents $67,897,500 (row 7, 36 percent bracket column) of national taxable income. Similarly $290,200-$250,00 or $40,200 of section 151(d)(3) scope is within the 39.6 percent bracket, which represents amount in Row 7, 39.6 percent column of national taxable income. Since the 39.6 percent tax bracket has no ending and no width, the 36 percent bracket figures for national taxable income per dollar of bracket width had to be used for attributing national taxable income to 39.6 percent bracket phase-out amounts.
/19/ Repeal of the phaseout taxes will mean that taxable income shrinks as exemptions become fully available, which will drop some income into lower tax brackets. Section 151(d)(3) is responsible for 5.79 percent of taxable income (row 3) in the bottom of the new 40.6 percent bracket, which will be shifted downward to the new 37 percent bracket. On average that will drop half of 5.79 percent or 2.9 percent of national taxable income within the bracket to the lower bracket, dropping revenue from 40.6 percent to 37 percent or 3.6 percent. Row 16 represents 3.6% * 2.9% * row 15 figures. There is no assumed similar shift from the 37 percent bracket into the 31 percent bracket, because the starting point of the section 151(d)(3) phaseout ($167,700) is sufficiently high within the 36 percent bracket ($140,000 start) to prevent the shift.