Senate Leadership Tax Proposals:
Mushrooming Tax Cuts For High-Income Taxpayers
Would Jeopardize Long-Term Budget Integrity

by Iris J. Lav

Summary

On January 22, 1997, the Senate Republican leadership introduced S. 1 and S. 2 — its tax cut proposals for the 105th Congress.¹ These bills are striking in several respects. While they include most of the Contract with America tax cuts contained in the budget reconciliation bill that Congress passed in November 1995 (and President Clinton subsequently vetoed), they go beyond the 1995 bill in two key respects. First, over the long term, the new tax cut proposals are more expensive than those in the 1995 legislation. Second, the new tax cuts provide a larger share of their tax benefits to high-income individuals than did the 1995 legislation, which itself favored such individuals heavily.

• The Congressional Joint Committee on Taxation, Congress' official scorekeeper on tax legislation, has issued estimates of the costs of the new bills. The estimates show the Republican leadership's two tax bills would cost $201 billion in the five-year period through fiscal year 2002.² S.2, the principal tax bill, costs $193.4 billion during this period, while S.1 — which includes tax breaks for higher education costs — costs $7.1 billion.

• But looking only at the five-year figures leaves misleading impressions. These tax proposals are heavily backloaded; they are designed in such a way that their full cost is masked during the first five years and grows rapidly after that. It also should be noted that the five-year cost figures for the new bills should not be compared to six-year or seven-year cost figures for tax bills Congress passed or considered in 1995 or 1996.

• The 10-year cost estimates the Joint Tax Committee has issued provide a more accurate picture of the full costs of these proposals when they are fully in effect. The Joint Tax Committee estimates show the 10-year cost of S.1 and S.2 is $526 billion. The cost during the second five years — $325 billion — is 60 percent greater than the $201 billion cost during the first five years.

• Moreover, the cost reaches $73 billion a year in 2007, the tenth year. It grows still larger in years after that. In the second 10-year period the legislation would be in effect — the years from 2008 through 2017 — the legislation would cost approximately $760 billion.

• When the cost of the tax cut reaches $73 billion in fiscal year 2007, the program cuts that would be needed to balance the budget would be 50 percent deeper than would be the case without tax cuts. In 2002 when the tax cuts total $53 billion, the program cuts required to balance the budget would be 39 percent deeper than would otherwise be the case.

• The Leadership tax proposals have been portrayed in some news accounts as less costly and more moderate than prior Republican tax proposals. Such portrayals are not accurate. Over the next 10 years, the Leadership proposals cost $24 billion more than the tax cuts in the vetoed 1995 reconciliation bill. (While the leadership tax cuts cost $20 billion less over the first five years than the tax cuts in the vetoed bill, they cost $44 billion more over the subsequent five years than the vetoed tax cuts.³) Over the long term, the new proposals cost significantly more than those in the vetoed legislation.

The Leadership proposal costs more than the tax cuts in the 1995 reconciliation bill primarily because the capital gains, Individual Retirement Accounts and estate tax breaks in the new plan are substantially larger and more costly than those proposed in 1995. Over the next 10 years, these three provisions — all of which primarily benefit upper income households — would cost $120 billion, or 85 percent, more than they would have cost under the first 10 years of the 1995 legislation. (The total package costs only $24 billion more than the 1995 bill because it does not include a number of smaller tax reductions the 1995 legislation contained.)

• The contrast is even greater when the new Leadership tax proposals are compared to those assumed in the budget resolution the Republican Congressional majority adopted in June 1996. The budget resolution passed last year included $122 billion in tax cuts over six years, compared to the $201 billion the new Leadership proposals would cost over five years.

• With their $201 billion cost over the initial five years, the Leadership tax proposals are sometimes described as being double the $109 billion cost of the Administration's tax cut proposals. In fact, the differences are much larger than that. The $109 billion figure reflects the gross cost of the Administration's tax cut proposals before the Administration's $66 billion in offsetting revenue-raising measures are taken into account; the net cost of the Clinton tax proposals is $43 billion over the next five years.⁴

Even if one assumes the Republican leadership would support all extensions of expired excise taxes the Administration has proposed — and if the savings from extending these expired taxes are netted against the cost of the Leadership plan — the net cost of the Republican leadership proposal over ten years is still more than three times larger than the net cost of the Clinton proposal. Measured in this fashion, the Republican plan would cost $440 billion over ten years, while the Clinton plan would cost $131 billion.

• Nevertheless, the Administration tax proposal also is substantially backloaded; it grows rapidly in cost after 2002. The Clinton proposal, net of offsets, would cost twice as much in 2007 as in 2002 — nearly $21 billion in 2007 compared to just over $10 billion in 2002.

• The principal Leadership tax bill, S.2, is called the "American Family Tax Relief Act" and is presented as primarily benefitting middle-class families. When the tax cuts take full effect, however, the primary beneficiaries are high-income households, not those in the middle class.

Using a generous reading of what constitutes middle-income tax relief, 61 percent of the tax cuts in the Senate Leadership package go to middle-class families during the first five years. But in the second five years, only 33 percent of the tax cuts go to such families. By the tenth year, less than 30 percent of the tax package is devoted to middle-income tax relief. This occurs because those tax cuts that are heavily backloaded in cost — and that eventually constitute the bulk of the tax package — are provisions primarily benefitting high-income households. The true dimensions of the tax benefits for the well-off are masked during the first five years; they become apparent only after that.

• The Senate leadership tax cuts are more heavily skewed to high-income people than those in the vetoed 1995 conference agreement. In the tenth year of the 1995 bill, tax relief provisions that primarily benefit middle-class families or small businesses accounted for 46 percent of the tax cuts in that legislation. But under the current Senate leadership proposal, middle-income tax relief makes up 30 percent of the tax cut benefits in the tenth year.

• A recent analysis by Citizens for Tax Justice, using the Microsimulation Tax Model of the Institute on Taxation and Economic Policy, found that only one-quarter of the tax cut benefits in S. 2 would accrue to the bottom 80 percent of U.S. households when the tax cuts are in full effect. The analysis also found that more than half of the benefits would go to the top five percent of households.⁵

• The new proposals would abrogate the quid pro quo that formed the framework of the landmark Tax Reform Act of 1986 — that in return for sharp reductions in the tax rates that high-income individuals faced, the tax preferences they enjoyed in the capital gains, IRA, and other areas would be narrowed. Under the Leadership tax package not only would income tax rates on the wealthiest Americans remain far below the levels in place before enactment of the 1986 Tax Reform Act, but capital gains, IRA, and estate tax breaks would be even more lucrative for such individuals than they were prior to 1986. The capital gains tax would shrink to its lowest level in 40 years.

• Finally, because most states incorporate the federal definition of "adjusted gross income" into their state income tax codes, enactment of the capital gains and IRA proposals at the federal level also would result in state tax cuts that cause states to lose billions of dollars in revenue. The federal estate tax reductions would result in state revenue losses as well. Here, also, the principal beneficiaries of the tax cuts would be individuals at high income levels.

These large tax cuts are being proposed at a time when there appears to be growing bipartisan agreement on the desirability of reaching budget balance in coming years. Moreover, some economists and budget experts are recommending the government run modest budget surpluses in the years before the baby boom generation retires to boost national saving and strengthen economic growth, so we ultimately will be better able to afford the large retirement costs of that generation. The larger the long-term cost of new tax cuts, however, the more difficult it will be to attain such goals and the more adverse the nation's long-term fiscal outlook is likely will be. Tax cuts of the magnitude the Senate Leadership has proposed eventually must result in either bigger deficits or cuts in basic programs serving millions of Americans that are much larger than either party currently admits favoring.

I. The Senate Republican Leadership Tax Plan

On January 22, 1997 the Senate Republican leadership introduced two bills that constitute its tax cut proposals for the 105th Congress. The principal tax bill, S. 2, has four components. It includes:

• a tax credit of up to $500 per child for middle-income and some upper-income families;

• a sharp reduction in the taxes on capital gains income earned when assets such as stocks, bonds, or real estate are sold;

• a repeal of the income limits on deductible contributions to Individual Retirement Accounts by taxpayers who are covered by an employer-sponsored retirement plan, along with the creation of a new, backloaded type of IRA that could be used by all taxpayers; and

• an increase from $600,000 to $1 million -- and in many cases well over $1 million -- in the amount of an estate that can be passed to heirs without any taxes being paid.

According to the Joint Committee on Taxation, the proposals in S. 2 cost $193.4 billion over the next five years and $507.8 billion over the next 10 years.

The other bill in the leadership package, S. 1, contains proposals that provide tax breaks both for certain post-secondary education expenses and for savings to help pay for future higher education costs. S. 1 also would extend permanently a provision of current law allowing taxpayers to exclude the value of employer-provided educational assistance from their income. The Joint Committee on Taxation has estimated the cost of S. 1 to be $7.1 billion over the next five years and $18.0 billion over the next 10 years. ⁶

This report examines the costs and distribution of the proposed tax provisions in these bills and how these proposals compare to earlier Republican Congressional proposals. A final section of the report looks at the implications of the package for efforts to eliminate the federal deficit and maintain balance after 2002. An appendix examines the Administration tax proposals and compares them to the Senate Republican leadership plan.

II. The Child Tax Credit and "Middle Class" Provisions

Over the first few years, the child tax credit is the most costly component of the Leadership tax package. By 2007, however, the child credit dwindles to less than a quarter of the cost of the total tax package.

The Leadership credit is identical to the child tax credit included in the vetoed November 1995 conference agreement. It provides a credit of up to $500 per child for middle-income families — and for many upper-income families as well — that can be used to offset their income tax liability.

Low-income working families that owe no income tax would not be eligible for the credit. Moderate-income working families that owe modest amounts of income tax would receive only a partial credit, irrespective of the amount of payroll taxes and federal excise taxes they pay. Some 31 million children — 44 percent of all children — live in low-and moderate-income families that would receive no credit or only a partial credit.

At the other end of the income spectrum, the credit begins to phase out for married couples filing jointly who have incomes above $110,000 and for heads of household with incomes above $75,000. A couple with two children would become ineligible for the credit once the couple's income reached $150,000. Only five percent of children live in families with incomes too high to qualify for a full credit, as compared to the 44 percent of children who live in families with incomes too low to receive a full credit.⁷

The cost of the child tax credit declines over time. This stands in contrast to the tax cut provisions that primarily benefit high-income households, which swell in cost over time. The cost of the child credit declines over time largely because the child tax credit is not indexed for inflation. The child tax credit costs $21.3 billion in 1998, the first year it is in full effect; by 2007, its cost is $16.9 billion. Similarly, its cost drops from $109 billion in the first five years to $89.9 billion in the second five years.

The lack of inflation indexing for the child credit contrasts with the legislation's treatment of capital gains. S. 2 would index capital gains for inflation, a measure that primarily benefits high-income households, while declining to index the child credit.

Tax Provisions Geared to Middle-Class Families

The child tax credit is the major piece of the Leadership tax package that primarily benefits middle-income taxpayers. As noted, even this proposal favors those in the top half of the income scale over those in the bottom half, since most children in low- or moderate-income families either are ineligible for the credit or qualify for only a partial credit. Moreover, a recent analysis by Citizens for Tax Justice, using the tax model described earlier, found the bottom 60 percent of tax filing units would receive only 25 percent of the tax cut benefits provided by the child tax credit proposal, while the top 20 percent of tax filing units would garner 37 percent of the benefits. Even so, we classify this provision as a middle-income relief provision in this analysis because the top fifth of households would not secure a majority of the tax cut benefits. The provisions we classify as upper-income tax relief proposals would confer a majority of their benefits on the top fifth, and in some cases, on the wealthiest five or 10 percent of households.

Two other components of the Leadership proposal also arguably provide the majority of their benefits to middle-income families. They are: the provision allowing a withdrawal of funds from Individual Retirement Accounts free of tax or penalty for such purposes as a business start-up, long-term unemployment, or college education; and the provisions in S. 1 that provide tax incentives to households for certain types of education savings and expenditures.⁸ Taken together, the tax relief provided by these three middle-income relief provisions accounts for 61 percent of the cost of the Leadership tax bills in the first five years.

The picture changes dramatically, however, in the second five years. By then, the three tax cut provisions providing a majority of their benefits to middle-class households constitute only 33 percent of the overall tax cut costs of the legislation. By 2007, these provisions account for just 29 percent of the tax cuts in the leadership package.

In short, after the first few years, the proportion of the tax cuts going to middle-class families declines sharply while the proportion going to high-income families and individuals shoots up. The erosion in the value of the child tax credit accounts for only a modest share of this shift. The principal reason the tax cut package is so much more heavily geared to high-income households after the first few years is that the cost of the legislation's upper-income tax breaks is kept artificially small in the early years but then mushrooms.

The principal provisions of these tax bills that primarily benefit high-income households — the capital gains, Individual Retirement Accounts, and estate tax provisions — all contain devices whose principal purpose is to mask or reduce the cost of these tax breaks in the period through 2002. These devices succeed in holding down the costs of these provisions over the next five years. But the costs of the provisions grow rapidly after that.

III. The Capital Gains Provisions

The Leadership proposal would substantially reduce the taxation of capital gains income — that is, income from the sale of assets such as stocks, bonds, and real estate. For individuals with capital gains income, the bill proposes two changes.

First, half of the profits from the sale of these assets would be excluded from taxation. This change alone would bring the maximum effective tax rate on capital gains income down from its current level of 28 percent to 19.8 percent. That would constitute the lowest tax rate on profits from the sale of such assets in more than 40 years.

The second change would go still further — it would assure that most capital gains would be taxed at effective rates significantly below the 19.8 percent rate. Under this change, taxpayers who sell assets held for at least three years would be allowed to exclude from taxation the portion of their profits that can be said to reflect the effects of inflation

Through the combination of a provision excluding half of the profits from the sale of these assets from taxation and the inflation indexing provision, the Senate leadership proposal would allow a substantial majority of capital gains income to escape any taxation. Consider a stock initially purchased for $100,000 and sold after five years for $134,000. Under current law, the investor would pay tax at a maximum rate of 28 percent on the $34,000 capital gain; the tax would equal $9,520. Under the proposed legislation, however, the price for which the investor purchased the stock would be adjusted for inflation and then only half the remaining gain would be subject to taxation. Assuming inflation of three percent per year, the $100,000 purchase price first would be adjusted upward to approximately $116,000 to account for five years of inflation; the investor's capital gain thus would be considered to be $18,000 rather than $34,000. Then 50 percent of this $18,000 gain would be excluded, making only $9,000 of it taxable. If the taxpayer was in the top income tax bracket (the 39.6 percent bracket), the taxpayer would pay tax of just $3,580. This represents a 62 percent tax reduction from the $9,520 in capital gains tax the investor would pay under current law.

Looked at another way, under current law, the investor would pay tax equal to 28 percent of the $34,000 profit. Under the Leadership proposal, the investor would pay tax of only $3,580 on this gain, an amount equal to just 10.5 percent of the $34,000 profit. This wealthy investor would pay a lower effective rate of tax on this profit than moderate-income families must pay on their wages and on interest they receive on modest savings accounts.

Table 1 shows the dimensions of the proposed capital gains tax cut under several scenarios. In each case, wealthy investors in the top tax bracket would pay a lower rate of tax on capital gains income they derive from buying and selling stocks, bonds, and other assets than modest-income families would pay on wages and interest earnings.

This capital gains indexing provision is one of the key provisions whose costs are made to appear artificially low in the first few years, masking its long-term drain on the Treasury. The initial costs are kept low through a device that causes the capital gains indexing provision to save, rather than cost, money on a one-time-only basis in 1998.

Under the indexing provision, the price an investor pays for an asset purchased after December 31, 1996 would be adjusted upward for inflation when the investor sells the asset. Investors would, however, be allowed to consider assets they already hold on December 31, 1996 as assets that they have newly purchased — and that consequently qualify for inflation indexing after that date — if the investors pay tax on the increase in the value of such assets between the time the assets were actually purchased and January 1, 1997.⁹

Suppose an investor has an asset he originally purchased for $100,000 and that is worth $120,000 on January 1, 1997. The investor would pay capital gains tax on the $20,000 gain in the value of the asset from the time of its purchase to January 1, 1997. (Because the legislation excludes half of capital gains profits from taxation, $10,000 of this $20,000 gain would be excluded, so the investor actually would pay tax on $10,000 in profit.) From January 1, 1997 on, the purchase price of the asset would be considered to be $120,000, and this $120,000 figure would be adjusted for inflation from that time forward.

The payment of taxes on the increase in the value of assets between the time of their purchase and January 1, 1997 would yield revenues of approximately $11 billion in fiscal years 1997 and 1998, when tax payments for calendar year 1997 are due. Most of these revenues would represent taxes the federal government would have collected anyway in subsequent years when the assets were sold. As a result, most of the revenue gain in 1997 and 1998 would represent a "timing shift" rather than a true increase in government receipts. Over the long term, capital gains indexing causes very large revenue losses.

Proposals to reduce capital gains taxes also are assumed to bring in a boomlet of additional revenues shortly after enactment, because some taxpayers who would hold their assets at the higher tax rate presumably would be willing to sell them at a lower tax rate. Together, the "mark-to-market" procedure used to qualify assets for indexing and the additional asset sales that would occur as a result of a lower capital gains tax rate would result in the proposed capital gains tax cuts for individuals raising approximately $14 billion in revenue in fiscal years 1997 and 1998. Once again, most of the added revenue collected in these years would represent revenue that otherwise would have been collected in future years.

After 1998, the capital gains provisions begin to lose large sums. The Joint Tax Committee estimates show that in fiscal year 1999, the capital gains tax cuts for individuals would lose $4.5 billion. By 2002, the loss would be nearly $14 billion. By 2007, the loss reaches $20 billion and is continuing to grow.

S. 2 also contains a cut in taxes on corporate capital gains — that is, on profits corporations make when they sell stocks, bonds, or other assets. S. 2 reduces the maximum tax rate on corporate capital gains income from 35 percent under current law to 28 percent. In other words, the bill cuts corporate capital gains taxes by one-fifth.

When the individual and corporate capital gains tax cuts in S. 2 are considered together, the cost is $33 billion during the first five years. Over the second five years, however, the cost is $96 billion — nearly three times as much. The total cost is $129 billion over the first decade.

In 2007, the cost of these capital gains provisions reaches $22 billion a year and is still rising. In subsequent 10-year periods, the cost of the capital gains tax cuts in S. 2 would exceed $200 billion.

Who Benefits from the Capital Gains Tax Cuts?

Many middle-income families have capital gains income, but the amounts of such income they receive are typically small. High-income households are more likely than middle-class households to have investments yielding capital gains income. More important, the average amount of the capital gains income that high-income households receive greatly exceeds the amount that middle-income investors typically receive. Studies generally show that households with incomes in excess of $100,000 receive approximately three-quarters of all capital gains income.

As a result, it is not surprising that Joint Tax Committee studies of earlier capital gains tax cut proposals found them to be heavily skewed toward the well-to-do. For example, when the Joint Tax Committee studied President Bush's 1992 proposal to exclude from taxation 45 percent of capital gains income received from the sale of certain assets, it found that 70 percent of the benefit from this tax cut would go to households with incomes of $100,000 or more. The Joint Tax Committee also found that households with incomes exceeding $200,000 would capture more than half of the tax cut benefits from that proposal. A separate Joint Tax Committee study of capital gains indexing found that half of the benefits from indexing would accrue to households with incomes of $200,000 or more.¹⁰

The Citizens for Tax Justice analysis of the distributional effects of the Senate leadership tax proposal (described above) yielded similar results. It found that the more than 64 percent of the benefits of the capital gains provisions in the Leadership plan would go to the top one percent of households, those with incomes exceeding $240,000.

IV. The Individual Retirement Account Proposals

The Senate Republican Leadership proposal would repeal all income limits on deductible IRA contributions by taxpayers already covered by an employer-sponsored retirement plan. Currently, individuals who are covered by an employer plan may make tax deductible contributions to Individual Retirement Accounts if their incomes do not exceed $35,000 for an individual or $50,000 for a couple. For those without an employer-sponsored retirement plan, there is no income limit.

Before the Tax Reform Act of 1986, taxpayers with incomes above the $35,000 and $50,000 limits who were covered by an employer retirement plan also could make deductible contributions to IRAs. The removal of the tax deduction for IRA contributions by such people was one of the changes the 1986 Tax Reform Act made in exchange for the Act's sharp reduction in the tax rates that upper-middle and higher-income households pay.

The Republican leadership tax plan would gradually lift the $35,000 and $50,000 income limits on IRA deductions to $135,000 for individuals and $150,000 for couples by 2000 and then eliminate the income limits altogether in 2001. Beginning in 2001, taxpayers at all income levels — including upper-income taxpayers with tax-favored private pension coverage — would be eligible to make deductible contributions to IRAs.

Under current law, a person eligible to make a deductible IRA contribution also may make an additional contribution of up to $2,000 on behalf of a non-working spouse, if the individual or the individual's spouse is an active participant in an employer- sponsored retirement plan, the income limits described above apply. By contrast, under the Senate Republican leadership proposal, an individual covered by an employer-sponsored retirement plan could make deductible IRA contributions for himself or herself as well as for a non-working spouse without regard to the income limits in current law.

In addition to these changes in the eligibility rules for deductible IRA contributions, the legislation would make all taxpayers eligible for a new type of Individual Retirement Account tax break called an "IRA Plus." Under this new type of IRA account — often called a "backloaded" IRA account — contributions to the account would not be tax deductible. But all interest earned on funds in the account would be permanently free of tax.

For taxpayers who participate in other types of tax-deferred savings plans, such as a 401(k) plan, the total of contributions to a deductible IRA and to another tax-deferred savings plan could not exceed the annual limit on voluntary contributions to the tax-deferred plan, which is generally $9,500. But all taxpayers, including those making the maximum contributions allowed to tax-deferred savings plans, would be eligible to make the maximum contributions to an IRA Plus account on top of their contribution to the other plans. Thus, a high-income taxpayer could contribute $9,500 to a 401(k) plan, another $2,000 to an IRA Plus account, and still another $2,000 to an IRA Plus account on behalf of a non-working spouse, and receive tax breaks on all of these deposits.¹¹

A separate provision in S. 1 would allow taxpayers to make an additional non-deductible annual deposit of $1,000 for each child under 18 to a "Bob Dole Education Investment Account." This account would operate like an IRA Plus; earnings on the account would accumulate free of tax, and no tax would be owed on amounts later withdrawn for higher education expenses.

Thus, under the Republican Leadership proposal, a high-income taxpayer already deferring tax on $9,500 a year in contributions to a tax-deferred savings plan could contribute an additional $2,000 each year to an IRA Plus, another $2,000 to an IRA Plus on behalf of a non-working spouse, and still another $1,000 a year for each child to an education account. As the years passed and such deposits accumulated — with many taxpayers shifting additional funds each year from other saving or investment vehicles to these accounts — the interest on a growing share of saving in the United States would be permanently sheltered from taxation.

This is a larger expansion of IRAs than was proposed in the vetoed 1995 reconciliation bill. In that bill, the income limits for deductible contributions to IRAs by taxpayers covered by employer-sponsored plans were not eliminated. Instead, the income limits for deductible IRAs gradually were increased, reaching $95,000 for individuals and $120,000 for couples by 1997 and being adjusted for inflation after that.¹²

The Costs of the IRA Proposals

The revenue losses from these IRA provisions would mount rapidly over time. Revenue losses would total $32.6 billion during the first five years the proposals were in effect. But over the second five years, the revenue losses from these provisions would total $80 billion — two and one-half times as much as in the initial five years. (These figures do not include the costs of the Bob Dole Education Investment Accounts.)¹³

According to the Joint Tax Committee, the revenue loss from the IRA provisions is $19.4 billion in 2007 alone and is still growing by $1 billion to $2 billion a year at that time. As a result, the IRA provisions would cost at least $200 billion over subsequent 10-year periods. Since the capital gains tax cuts also would cost that much in subsequent 10-year periods, the combined loss from the capital gains and IRA provisions would exceed $400 billion in subsequent 10-year periods.

The IRA provisions thus share a common characteristic with the capital gains provisions in that the revenue losses which both provisions produce start relatively small and grow markedly over time. The Leadership tax bills employ a series of gimmicks to keep the cost of the IRA provisions artificially low in the period between now and 2002, with the result that their full costs begin to emerge only after then.

Gimmicks Used to Backload IRA Costs

One of the principal devices used to keep the costs of the IRA provisions low in the initial years is the creation of backloaded IRAs. The concept of a backloaded IRA was developed a few years ago specifically to evade federal budget rules which require that the cost of a tax cut to be fully "paid for" over its first five years through either an offsetting tax increase or a reduction in an entitlement program. If the cost of a major IRA expansion can be kept artificially low during the first several years, that eliminates the need for a large, controversial tax increase or entitlement cut to pay for it.

Under a backloaded IRA, there is little revenue loss in the initial years because no upfront $2,000 tax deduction is allowed for IRA deposits. Instead, the large revenue losses occur when depositors retire and begin to withdraw funds from these IRA accounts without having to pay any tax on them whatsoever (because all interest earned on backloaded IRA accounts is permanently tax free). Most of the revenue losses from backloaded IRAs thus are conveniently moved outside the five-year budget period that ends in 2002.

The Leadership proposal also includes a second device whose purpose is to keep revenue losses low in the initial years. The proposal would permit taxpayers to take funds they currently have in conventional IRA accounts and transfer these funds, with no penalty for early withdrawal, to backloaded IRAs on which all future interest earnings would be tax free.

At the time of transfer, the funds moved from a conventional IRA to a backloaded IRA would be taxed as ordinary income. But if the transfer were made before 1999, the income could be averaged over a four-year period. Averaging the IRA income in this manner would lower the tax that many taxpayers would owe on it, because averaging would keep the IRA income from pushing many of these taxpayers into a higher tax bracket. This four-year averaging period conveniently spans the years from 1999 through 2002.

The purpose of this provision — known as a "rollover" provision — is to generate a one-time boost in revenues in the next few years as depositors move money from old IRAs to backloaded IRAs and pay tax on the funds they transfer. (Many IRA account-holders would make these transfers so that future earnings on their IRA accounts would be tax free.) The taxes collected now as these funds are transferred from one type of IRA to another are taxes that would have been collected in future years when the IRA account-holders retired. The result thus is to shift into the period through 2002 many billions of dollars in revenue that otherwise would be collected in later years.

The effect of these various IRA gimmicks is two-fold:

• The backloaded IRAs shift into years after 2002 many billions of dollars in revenue losses resulting from making upper-middle-income and upper-income individuals who already have tax-favored retirement plans eligible for generous IRA tax breaks. Because of this backloading gimmick, little of the large revenue loss from making these individuals eligible for IRA tax advantages shows up until after 2002.

• Meanwhile, the "rollover" gimmick accelerates into the years between now and 2002 many billions of dollars of tax revenue that otherwise would be collected in years after that. These billions of dollars offset much of the remaining cost of the IRA expansion over the next five years. This makes the cost of the proposed IRA tax breaks appear deceptively modest when only the cost through 2002 is examined.

While the proposed IRA provisions in S. 2 lose an average of $6 billion a year in the first five years, they lose over $19 billion a year by 2007 and more in years after that.

Who Benefits From These IRA Proposals?

The IRA changes in the Leadership tax bills would primarily benefit high-income families that also receive tax advantages through retirement plans their employers sponsor. These IRA proposals would not benefit most middle- and lower-income taxpayers because most such taxpayers already are eligible for IRA tax deductions under current law.

• A substantial majority of taxpayers already qualify for IRA tax deductions; in 1995, more than 70 percent of all taxpayers with earnings were eligible to make deductible IRA contributions. (Some 56 percent of all taxpayers with earnings are eligible to make the full maximum deductible contribution, while another 15 percent of taxpayers were eligible to make a partial contribution.) Since these taxpayers already can get these tax deductions, they would derive either no benefit or only a small benefit from the Leadership IRA proposals.¹⁴

• In addition, IRS data show that in 1986, the last time all taxpayers were permitted to make deductible contributions to IRAs regardless of their income level or pension coverage, 82 percent of IRA deductions were taken by the one-third of tax filers with the highest incomes.

Moreover, the IRA deductions these high-income individuals received in 1986 were worth more than the IRA deductions taken by taxpayers at lower income levels, since the high-income IRA users were in higher tax brackets. As a result, the top one-third of tax filers received substantially more than 82 percent of the IRA tax benefits.

For these reasons, a Joint Tax Committee analysis of an earlier proposal to restore eligibility for IRA tax deductions for individuals who have other tax-favored retirement plans and incomes above the current IRA income limits found that the top one-fifth of taxpayers would receive approximately 95 percent of the tax benefits from the proposal. The Joint Tax Committee also found that the most affluent five percent of taxpayers would collect nearly one-third of the tax benefits from a similar IRA proposal. The new IRA proposals are likely to have similar effects.¹⁵

V. Proposed Cuts in Estate Taxes

The Leadership proposal also cuts estate taxes. The proposal would increase substantially the amount of an estate that can be passed on to heirs without any tax being paid.

Under current law, the first $600,000 of an estate is exempt from tax. The Leadership proposal would gradually raise the amount of an estate that heirs can receive free of tax to $1,000,000 by 2004.

The proposal also includes an expanded estate tax exemption for qualified family-owned business interests. In addition to the $1 million exemption just mentioned, heirs could exclude $1.5 million of a family-owned business interest plus 50 percent of the remaining value of the business interest. These additional exclusions would apply when family-owned business interests comprise more than 50 percent of an estate.

Thus, if a decedent leaves an estate of $3 million, and $2 million of it consists of family-owned business interests, estate taxes would be owed on only $250,000 — or one twelfth — of the estate. Some $1 million of the estate would be excluded under the general credit, while $1.5 million in business interests, plus 50 percent of the remaining $500,000 in business interests, also would be excluded.

In this instance, $2.75 million of the $3 million estate would be shielded from tax. If the estate consisted of $2.5 million, with $1.5 million of it in family-owned business interests, the entire estate would be tax free.

The proposed exclusion for family-owned business interests is a new tax break. Under current law, there is no special estate tax exclusion for qualified family-owned business interests, although the $600,000 exemption can be applied to such businesses.¹⁶

Like the capital gains and IRA tax cuts, the cost of the Leadership proposal to cut taxes on large estates increases markedly over time. As noted above, the estate tax cuts would phase in between now and 2004. The Joint Tax Committee estimates show the estate tax provisions would cost $18 billion in the first five years but $48 billion in the second five years. By 2007, these provisions would cost nearly $12 billion a year. This suggests their cost in subsequent 10-year periods is likely to exceed $120 billion.

Only one percent of estates exceed $600,000 in value. Thus, these provisions would have no effect on the 99 percent of estates valued at less than $600,000 and would not benefit ordinary middle-class families.17 The provisions would reduce or eliminate tax only on estates valued between $600,000 and $21 million. The benefits of the exclusion would be recaptured fully for estates valued above $21 million. 18

VI. Long-term Cost of the Leadership Proposals

As this description of the specific provisions of the Leadership tax proposal indicates, the large tax cuts primarily benefitting higher-income taxpayers would grow rapidly in cost over time and cost far more after 2002 than in the years before it. By contrast, the tax benefits the child tax credit confers would erode as the years go by.

Table 2 summarizes the Joint Tax Committee estimates of the cost of the Leadership tax proposals over the first 10 years. The table also estimates the cost of these provisions over the subsequent 10-year period.

Table 2
Cost of the Leadership Tax Proposals
(in billions of dollars)

As the table shows, the cost of the Leadership tax package rises from $201 billion in the first five years to $325 billion in the second five years. By 2007, the cost is $73 billion in a single year.

By examining the cost of these proposals in the last several years of this 10-year period, it is possible to estimate their cost in the subsequent 10-year period as well. Over the 10-year period from 2008 to 2017 — a period when the baby boom generation will begin to retire, placing substantial added burdens on the Treasury — the cost of the Leadership tax cuts rises to approximately $760 billion.

These figures reflect the fact that the Leadership proposals continue to grow in cost after the first 10 years. Consider the capital gains indexing provision as an example. Under this proposal, the number of years over which the value of an asset is adjusted for inflation would increase over time. In 2007, the value of an asset could be indexed for a maximum of 10 years. By 2017, a maximum of 20 years of inflation could be taken into account. Similarly, by 2017 the amount of deposits in IRA Plus accounts — on which all earnings would permanently be exempt from taxation — would be much larger than the amount of such deposits in 2007.

VII. Comparing the Cost of the New Proposals
to the Cost of the Vetoed 1995 Bill

Over the next 10 years, the Leadership tax proposals cost $24 billion more than the "Contract with America" tax provisions in the 1995 budget reconciliation package that President Clinton vetoed. Both plans include a child tax credit, a capital gains tax reduction, expansion of IRAs, and a reduction in estate taxes. But while the child tax credit is essentially the same in both plans, the capital gains, IRA, and estate tax breaks all are more costly in the new Leadership plan than in the 1995 legislation.

• The capital gains provision in the current proposal is more expensive in part because indexing would begin in 1997, whereas it would have started in 2001 under the vetoed bill. Under the 1995 legislation, the profit on an asset sold in 2001 would have been taxed at a maximum capital gains tax rate of 19.8 percent; indexing could not have been used to lower the effective tax rate to a lower level. By contrast, under the new proposal, the profit from the sale of such an asset would be adjusted downward for tax purposes by factoring in four years of inflation indexing. As a result, the profit from the sale of the asset would be taxed at an effective rate of about 10 percent or 11 percent, rather than 19.8 percent. Because the new proposal makes indexing effective for four additional years, it reduces the capital gains taxes on assets purchased before 2001 to a significantly greater degree than the 1995 legislation would have done. For this reason, the new bill has a larger and more costly capital gains tax cut than the earlier legislation did.

• In addition, the IRA provision in the new Senate Republican leadership plan is more costly than the IRA proposal in the 1995 bill. In the 1995 bill, the income limits on deductible contributions to IRAs would have been increased to $95,000 for individuals and $120,000 for couples by 1997 and adjusted for inflation after that. Under the new legislation, income limits on deductible contributions to IRAs would be eliminated entirely.

• Finally, the new leadership tax plan would expand tax breaks for large estates more than was proposed in the 1995 bill. Among other differences, the 1995 proposal would have raised the amount of an estate that heirs can receive free of tax from $600,000 to $750,000; the current proposal would raise the tax-free amount to $1 million. The current proposal also provides a more generous exclusion for inherited family-owned business interests; it would exclude the first $1.5 million of family-owned business interests, while the 1995 plan would have excluded the first $1 million.

The cost of the current Senate Republican leadership plan exceeds the cost of the vetoed 1995 reconciliation conference agreement for these three tax breaks by nearly $120 billion over the first 10 years, or more than 85 percent. (See Table 3.) That the total difference in cost between the two plans over the 10 years is a more modest $24 billion is due largely to the fact that the 1995 plan included a number of additional forms of tax relief that are not part of the new proposal. The child tax credit and the relatively modest education tax breaks in S.1 are the only provisions the new plan contains in addition to the three large upper-income tax breaks.¹⁹

Table 3
Comparison of 10-Year Costs of Selected Tax Provisions
in 1995 Conference Agreement and Senate Leadership Plan
(in billions of dollars)

VIII. Proposed Tax Cuts Break the Quid Pro Quo
of the Tax Reform Act of 1986

The Tax Reform Act of 1986 sharply lowered income tax rates on upper-middle income and high-income individuals, as well as on corporations, in return for scaling back tax breaks provided for capital gains, IRAs, estate taxes and other items. Under the Senate Republican leadership proposal, that quid pro quo would be abrogated. The income tax rates faced by those in the top brackets would remain far below the rates such taxpayers encountered prior to the 1986 Act. But capital gains and IRA tax breaks would not only be restored but also would be made more generous than they were before 1986. In addition, new estate tax breaks would be created. The result would be windfalls for wealthy Americans.

• Prior to 1986, the top income tax rate was 50 percent. While the top rate was raised somewhat in 1990 and 1993 after being lowered in 1986, at 39.6 percent it remains substantially below the pre-1986 rate.

• Special treatment for capital gains income was among the preferences ended in 1986 in return for taxing income at lower rates. Prior to 1986, taxpayers in the top income bracket paid an effective capital gains tax rate of 20 percent. The 1986 tax act eliminated the exclusion for capital gains; capital gains income became subject to a 28 percent maximum tax rate, the rate that remains in effect today.

• The Senate Republican Leadership proposal would reduce the maximum tax rate on capital gains income to 19.8 percent, slightly below the pre-1986 level. In addition, because of the capital gains indexing provision, high-income taxpayers would be subject to effective tax rates of only 10 percent or 11 percent on much of their capital gains income, a rate about half that in effect before 1986.

• Thus, compared to the pre-1986 law, the Leadership plan would leave the wealthiest taxpayers with both a 21 percent reduction in income tax rates and as much as a 45 percent reduction in capital gains taxation.

IX. Impact of Leadership Tax Proposals
on Balancing the Budget

The increasing cost of the Leadership tax proposals over time would make it difficult for the government to plan prudently for the second decade of the 21st century and the decades following that. When the baby boom generation retires, the nation will face a daunting fiscal challenge. The costs of Medicare, Medicaid and Social Security will swell as the number of elderly individuals mounts. In addition, the ratio of workers to retirees will fall to a low level.

To help the nation better afford the large retirement costs of the baby boom generation, some budget experts recommend building a surplus in the unified budget in the intervening period during which the baby boomers will be in their peak earnings years and using those annual surpluses to increase the amount of national saving. Increasing the amount of national saving would increase the likelihood of stronger economic growth, which in turn would better enable the country to afford to support the baby boomers when they retire.

To pursue this approach, the current deficit in the non-Social Security portion of the federal budget would need to be reduced significantly or eliminated in coming years. That would enable some or all of the reserves now building in the Social Security trust funds to be used to increase national saving. (If the annual deficit in the non-Social Security portion of the budget is smaller than the annual surplus in the Social Security trust fund, the unified budget will be in surplus.)

Even without a large tax cut, achieving a sizable surplus in the unified budget is a daunting task. The proposed Leadership tax cut would likely make such a task virtually impossible.

• Under current law, the Congressional Budget Office projects a budget deficit of $188 billion in 2002. Assuming a steady path of deficit reduction between now and 2002, CBO estimates that approximately $137 billion in program cuts would be needed in 2002 to bring the budget into balance. (The remaining savings, under the CBO estimate, result from a "fiscal dividend" from slightly stronger economic growth and reduced interest payments on the debt.)

• The Leadership tax proposals would result in a revenue loss of $53 billion in 2002. To balance the budget thus would require that program reductions total $190 billion rather than $137 billion. The tax cuts would thereby enlarge by 39 percent the magnitude of the program cuts required to achieve budget balance.

• If budget balance is attained in 2002, further actions will be required to maintain that balance in subsequent years. CBO projects a budget deficit of $278 billion in 2007 under current law. Assuming the "fiscal dividend" and interest savings that CBO estimates would result from reaching balance in 2002 and maintaining balance through 2007, program spending in 2007 would have to be $147 billion lower than current projections to maintain budget balance in that year. Accommodating the Senate leadership's tax cuts of $73 billion in 2007 would raise the required program reductions from $147 billion to $220 billion in that year. As a result, the proposed tax cuts would increase the size of the program cuts otherwise required to balance the budget in 2007 by 50 percent. (See Table 4.)

Reductions of this size would be extremely difficult to achieve. Measures that sought to go further and place the unified budget in surplus so a substantial part of the Social Security surplus could be saved would essentially be unattainable.

By making it necessary to secure substantial additional program cuts to achieve and maintain budget balance, the Leadership tax proposals also would be likely to have one other effect. As a result of the tax cuts, any balanced budget agreement would have to contain deeper, more controversial cuts in basic programs, which in turn would likely make bipartisan budget agreements harder to reach.

Table 4
Program Spending Cut Required to Achieve and Maintain
A Balanced Budget with Senate Leadership Tax Cuts
(in billions of dollars)

X. State Revenue Reductions

Finally, the capital gains, IRA, and estate tax proposals would reduce state as well as federal taxes on individuals with higher incomes. The capital gains and IRA proposals would reduce state taxes because 36 states and the District of Columbia incorporate the federal definition of adjusted gross income (AGI) in their tax codes.

For instance, the Leadership proposal would exclude from the definition of AGI half of all capital gains income. Because of state conformity to federal tax rules, enactment of this proposal would cause the profits excluded from income under the federal tax code also to be excluded under state income tax codes. In addition, the reduction in the federal taxation of capital gains that would result from indexing the purchase price of assets for inflation would further reduce the amount of capital gains included as income in most states' tax systems. State revenues thus would decline as a result of these changes in the federal tax code.

Making upper-middle and high-income individuals eligible for IRA tax deductions and backloaded IRAs would further shrink state revenues. The IRA deductions would decrease the amount of income subject to state taxation; since the deducted amounts would not be part of adjusted gross income for federal income tax purposes, these amounts would not be part of AGI under many state tax systems as well. In addition, the new backloaded IRA accounts would shelter interest income from state income tax as well as federal tax.

Moreover, the estate tax in all states is tied closely to the federal taxation of estates. The federal government in essence "shares" its estate tax with the states by allowing taxpayers to take a credit for state estate taxes paid up to a specified percentage of the federal tax. The sharing mechanism works as follows.

Federal tax law determines the amount of federal estate tax that heirs to an estate must pay. States also may levy a tax on estates. If states levy their own estate tax by computing a percentage of the federal estate tax liability, special rules apply. Under these special rules, heirs may subtract the state taxes paid, dollar for dollar, from federal estate taxes that otherwise would be due, up to a specified maximum percentage of the federal taxes. In other words, the heirs have the same amount of overall tax liability whether or not the state levies an estate tax; the state tax does not change the total amount of estate taxes collected but rather allows the state to share in a portion of the federal revenue. Every state uses this type of state estate tax, which is commonly known as the "pick-up" tax.

Because state estate taxes must be levied as no more than a specified percentage of federal estate taxes to qualify as a "pick-up" tax and benefit from the special rules, reductions in federal estate taxes will automatically reduce state revenues. Thus, the Leadership plan would result in a substantial reduction in states revenues from estate taxes.

If these Leadership tax proposals are adopted, states will lose billions of dollars in revenue. States also will face the same problem the federal government is likely to experience as a result of these tax proposals — sharply escalating revenue losses in future years. To offset these revenue losses, states would likely be forced to cut programs, raise other taxes, or take both types of actions.

State spending cuts typically affect low- and moderate-income families the most. Moreover, tax increases in most states are likely to affect low- and middle-income taxpayers disproportionately, because virtually all state codes are regressive. As a consequence, those who would benefit least from the state tax cuts that would result from enactment of the Leadership proposals at the federal level would likely be the households affected most adversely by the actions states would take to make up for the resulting losses of state revenue.

Appendix

End Notes