Giving Tax Credit Where Credit Is Due

A ‘Universal Unified Child Credit’ that expands the EITC and cuts taxes for working families

by Robert Cherry and Max Sawicky

Although the U.S. economy is in the 109th month of a record-length recovery and the federal budget is in its best shape in history, nearly one in five American children are in families whose income is below the poverty line. Almost 10 million children lack health insurance coverage and Medicaid benefits (U.S. Census Bureau 2000a). An effective means of helping these children would be a major expansion of the Earned Income Tax Credit that restructures and combines some of the other tax benefits available to families with children.

Why is it important that these tax benefits are reexamined today? The nation’s experiment in welfare reform has reduced caseloads radically on the premise that welfare applicants ought to be moved quickly into the labor market, despite the fact that income from full-time, year-round work at low wages can fail to equal benefits formerly available under public assistance programs. A full-time, year-round worker earning the minimum wage will remain below the poverty line, even with but a single child.

Equality of opportunity is a value that is widely shared but inadequately served by the current system. Although no one disputes that family income has an important role in the life chances of children, political complications arise when efforts are made to improve these chances by providing additional income to needy families. As it stands, public policy discriminates against the children of parents who are not employed, or whose attachment to employment is irregular. These children are economically penalized because of the behavior of their parents.

One might expect such problems to disappear for the children of employed parents, but this is not the case. The federal personal income tax, while superior to the array of radical reforms proposed in recent years, is deficient in its treatment of children in families whose labor earnings are below the median income level.

In the current recovery, many jobs paying modest wages have been created. The question this report addresses is, what can the tax system do for families whose incomes depend on low wages? Some of the specific tax issues that need addressed include:

• Families near and below the poverty line are made poorer by payroll tax liability – between two and three million people are pushed below the poverty line by federal payroll taxes (U.S. Census Bureau 2000b).
• Tax benefits for children increase with family income.
• Due to the combined effects of the tax system and public benefit programs, families with low incomes can face very high marginal tax rates on their labor earnings.
• Households with children headed by a single parent can face significant income losses if the parent marries.

Despite the multi-trillion dollar budget surpluses, no tax proposal seriously addresses these problems, with political leaders showing no enthusiasm for an expansion of public spending to address poverty and inequality. Meanwhile, Congress has just enacted a tax cut for employed recipients of Social Security, thereby enlarging projected deficits in the trust fund and exacerbating tensions between beneficiaries and younger generations paying into the fund. In this light, perhaps political leaders will agree that some consideration for low- and moderate-income families with children is in order.

What is the Earned Income Tax Credit?
The Earned Income Tax Credit (EITC) is a credit against federal personal income tax liability. Unlike most other tax credits, the EITC is refundable, that is, if the credit exceeds tax liability, the taxpayer receives the difference in cash from the Internal Revenue Service.

The credit is available only to the employed, with benefits varying with the number of dependent children in a household. The credit amount rises with earnings according to a fixed percentage for an initial range of income called the phase-in range. The effect is the same as a wage subsidy – for every dollar earned, the credit adds 8, 34, or 40 cents depending on whether the family has zero, one, or two children.

Once a taxpayer’s wages push the benefit up to the legal maximum, the benefit stays constant for an additional range of income. We call this range the ‘plateau’ of the credit. Finally, at a certain point, the credit begins to decrease with additional income. This is called the “phase-out” range, with phase-out rates varying depending on the number of children.

The EITC program presently costs over $30 billion annually, making it the largest entitlement program in the federal budget, aside from health programs and Social Security (Smeeding et al. 1999).

Table 1 shows the basic parameters of the credit for tax year 1999 (the basis for all the calculations in this paper). For a family with two children, the phase-in rate is about 40%, as noted above, meaning that for each $100 of wages, the taxpayer is entitled to a credit of $40. The phase-in range extends to $9,500 annually, at which point the maximum credit amount is $3,816. For incomes between $9,500 and $12,500 – the “plateau” range – the credit remains constant. For incomes between $12,500 and $30,850, the credit phases out at a rate of approximately 21%.

Table 1 shows that the credit amounts for families with children are much higher than that for a taxpayer without children, and that the phase-in and phase-out rates differ for families with a given number of children, as well as between families with either one or two children.

A limited amount of non-labor income is permitted for recipients of the EITC, but, by and large, the end of the phase-out range reflects the income ceiling. Another important feature of the current EITC is that there are no added benefits for families with three or more children. While there are many other rules that are beyond the scope of this paper, this report focuses on how the basic EITC parameters can be altered to accomplish some broad social objectives.

Evidence on the effectiveness of the EITC
Research suggests that the EITC is more effective in moving families over the poverty line than any other government program (Greenstein and Shapiro 1998). Use of the EITC is high, especially compared to other income support programs. One estimate claims that over 85% of those eligible for the credit apply for it (Scholz 1994). Furthermore, none of the available evidence suggests that employers currently use the credit to justify lowering workers’ wages.

A major reason for the popularity of tax credits such as the EITC over traditional public assistance is their inherent work incentive. From 1984 to 1996, employment rates for single mothers with children increased significantly. Meyer and Rosenbaum (1999) attribute the bulk of this increase to repeated expansions of the EITC. Eissa and Liebman (1996) and Dickert, Houser, and Scholz (1995) also found the program to increase labor force participation.

Other tax benefits for families with children
The EITC is not the only benefit made available under the personal income tax to taxpayers with children, and comparing the EITC to some of the other tax benefits enjoyed by higher-income families points up some important inconsistencies.

Table 2 provides a list of selected tax benefits. First on the list is the dependent exemption, which is $2,750 for tax year 1999. The exemption is the same for all, but it is worth different amounts for different taxpayers. Since the personal income tax has increasing marginal rates, the higher one’s tax bracket, the more an exemption is worth. If one’s income is taxed at 15%, as in the first bracket, $100 of exemption is worth $15 in tax savings. If one is in the second 28% bracket, that same $100 of exemption is worth $28. The total exemption of $2,750 is worth $770 in tax savings for those in the 28% tax bracket, a point of reference that will be relevant to this report’s proposed reform of the EITC. [1]

For the wealthiest taxpayers in the top 39.6% bracket – well under 1% of all taxpayers – the exemption saves $1,089 per child in taxes. For those in the 15% bracket – the majority of taxpayers – the savings are $413 per child.

For low-income families, an extra disadvantage of the exemption is that it is not refundable. If, for instance, a married couple’s income is $7,000, they will owe no federal income tax because the standard deduction for a married couple is $7,200. At this income level, a child’s exemption becomes superfluous, hence worthless. Low income tax liability means low tax savings from deductions or non-refundable credits. Since society presumably values all children equally, a higher tax savings for children in more fortunate families doesn’t make much sense.

Another benefit for families with children is the recently enacted Child Credit of $500 per child. This does not vary with income or work arrangement, except in the sense that it is not refundable. Like the benefits of the dependent exemption, this credit’s amount cannot exceed tax liability. For families with three or more children, a refundable “Additional Child Credit” is also available.

The next benefit shown in Table 2 is the Child and Dependent Care Credit, which defrays a family’s expenditures on care of children and other dependents. Here again the value of the credit can be greater for families with moderate incomes than for many with low incomes. For one thing, the credit is not refundable. The credit also only defrays expenses for services that are purchased from a licensed day care provider. Parents that can’t afford to pay for child care of this sort can’t use the credit. No credit is available to offset the cost of a family member performing such duties. In this sense, to some extent, the credit discriminates against stay-at-home mothers.

Finally, there is a deduction of expenses for child care, up to $5,000 annually, for workers in firms that offer flexible spending or “cafeteria plans.” Under this arrangement, a worker’s employer deposits part of the worker’s gross wages into an account that is shielded from income tax withholding. Funds in the account may be used to reimburse the worker for child care expenses. These plans are more prevalent in firms with relatively highly paid employees, so there is an element of income discrimination in them. As for the dependent care credit, a family must be able to afford child care to have a chance at the credit. The benefit is not refundable, since it has no value if there is no tax liability from which to shield income.

How do these benefits stack up, compared to those available under the EITC? For a family with two children in the 28% bracket, tax savings consist of $1,540 from exemptions, $1,000 from the Child Credit, and as much as $1,400 for child care expenses, for a total of $3,840. Only a family receiving the maximum EITC benefit approaches this level of tax benefits.

For families with three or more children, the disparity widens. Unlike the EITC, three of these benefits (the dependent exemption, the Child Credit, and the Additional Child Credit) increase if the number of children increases beyond two.

Figure 1 illustrates the change in income tax benefits due solely to the dependent exemption and Child Credit for a married couple with two children. These benefits take the form of tax savings; as noted above, they are not refundable. (These benefits are calculated by defining tax savings as the difference in tax liabilities due solely to the availability of the dependent exemption and Child Credit. It is assumed that without these provisions, the taxpayer would retain the present standard deduction of $7,200 for a married couple filing a joint return.)

The benefits begin at an adjusted gross income of $12,700. Without exemptions and Child Credits, a taxpayer would owe 15 cents of tax for every dollar of income earned after the $12,700 level is reached. But because of the exemptions and credits, such an income incurs no tax. Thus benefits begin to rise at a rate of 15%, as income grows from $12,700 to $24,867, offsetting $1,825 in taxes that would otherwise be owed.

The first “step” on the left in the figure applies to most of those in the 15% income tax bracket. It reflects a constant level of benefits over the range of income it covers. The second step, going from left to right, reflects the increase in benefits for those in the 28% bracket. The increase stems from the fact that, with a rise in the marginal tax rate from 15% to 28%, the exemptions rise in value. For a taxpayer in the 15% bracket, two exemptions of $2,750 each ($5,500 total) are worth 15% of $5,500, or $825. For a taxpayer in the 28% bracket, the value rises to 28% of $5,500, or $1,540. The $1,000 worth of Child Credits do not change in value, so total benefits in the 28% bracket are $2,540.

The income range in Figure 1 includes all but about 5% of taxpayers; not shown are the 31%, 36%, and 39.6% brackets (Internal Revenue Service 1998). Under current law, in these higher brackets the dependent exemption and Child Credit phase out. The high-income phase-outs affect fewer than 2% of taxpayers (Barthold, Koerner, and Navratil 1998).

In 1996, 76% of tax filers were in the 15% bracket or lower (Internal Revenue Service 1998), so the restricted range used in the figure takes in a large majority of the population. Non-filers almost always have low or zero incomes. The 28% bracket for a couple begins at $43,050 in taxable income. For a couple with a child, the equivalent in adjusted gross income is $61,833 (taking the value of the $500 Child Credit into account). Such families can have a gross income that is higher still. For instance, adjusted gross income does not include deductible contributions to employer-paid pensions. Without a doubt, families in the 28% bracket are above median levels.

Figure 2 shows the combined benefits of the EITC and income tax fo
r a married couple with two children. The “steps” shown are the same as in Figure 1. The pyramid on the left is the EITC benefit schedule. The line above the right side of the pyramid is the sum of benefits for the EITC and individual income tax. This sum reflects the overlap between the EITC phase-out range, which ends at $30,850 for families with two or more children, and the range of income tax benefits for children, which begins at $12,700 for a married couple with two children. These benefits do not cancel each other out because the EITC is refundable. In other words, the exemption and Child Credit reduce positive tax liability, after which the EITC is available as a cash refund.

In this section we have noted the varied tax benefits for children available to families at different income levels. For families receiving close to the maximum EITC amount, their benefits could exceed those of families with higher income. But for many others, the reverse is the case for a multitude of reasons.

First, many low-income families do not receive the maximum EITC benefit. Second, combined benefits from both the EITC and the income tax can still fall below those for many families in the 28% bracket. Furthermore, the discussion here confines itself to the exemption and Child Credit and excludes the other benefits listed in Table 2, not to mention the plethora of additional tax breaks not available to low-income taxpayers in general. Finally, the disparity of tax-based child benefits for families with different incomes widens as the number of children in a family increases. For instance, the maximum EITC benefit of $3,816 does not change for larger families, but the tax benefits for a family of four children in the 28% bracket are $5,080.

Improving the EITC
Three major issues have dogged the EITC over its 25-year lifetime: (1) its “implicit marginal tax rate,” (2) its possible marriage penalty, and (3) the complexity of the filing process. Although these shortcomings are not unique to the EITC – they bedevil a wide variety of tax and public benefit programs – they should still be remedied in the spirit of equity.

Addressing the “implicit marginal tax rate”
The phase-out of benefits (as described above) can be likened to a tax on wages, since an extra dollar of earnings triggers some reduction in benefits. In other words, if earnings rise by a dollar, income rises by less than a dollar. The phase-out rate is analogous to a marginal tax rate. This problem applies to any sort of public benefit that is conditioned by the recipient’s income. If income is a factor in determining eligibility for the benefit, and possibly the size of the benefit as well, at some income level workers will face a reduction of benefits, eventually becoming ineligible. Thus, the act of increasing one’s income can have a negative consequence in terms of taxes paid. By the same token, the phase-in is a supplement to wages, one that would be expected to enhance the incentive to work.

For an example of the negative side of the work incentive, suppose a couple with two children is earning $12,500 and receiving the maximum benefit of $3,816. As the couple earns additional income, the credit begins to phase out. If labor earnings go up by $100, benefits fall by $21, so net income only goes up by $79. [2]

Some analysts have voiced concerns that the magnitude of the implicit tax can be an important disincentive to work for household heads and couples who are already working enough to earn at least $12,500 annually. If the extent of this “tax” does discourage some work, then lowering it could encourage greater work effort. Sixty percent of current EITC recipients – nine million families – are in the phase-out range (Liebman 1999).

How high can this tax be? For some low-wage workers, it can be quite high (Giannarelli and Steuerle 1995). Consider a family with two children. As Table 1 shows, the phase-out rate for such a family is 21%. In addition, the family might receive benefits under the food stamp program (FSP). FSP benefits are phased out as income rises, to the tune of 24%. Then there is the payroll tax. Including the portion paid by the employer – which most economists assume is borne by the worker – the total tax rate is 15.3%. Finally, there is the first bracket of the federal personal income tax, with a tax rate of 15%.

Insofar as the income ranges of these implicit taxes overlap, the combined marginal effect is magnified. This analysis doesn’t even take into consideration state income taxes, the Medicaid program, and housing benefits, each of which increase the marginal effects further at various points. There are also the private costs of working, such as child care, work uniforms, transportation, and the like. A basic goal of any reform should be to alleviate the pile-up of assorted tax and benefit programs at low-income levels.

The notion of the EITC creating a disincentive to work in certain circumstances is controversial because its magnitude is in doubt and subject to misinterpretation.

The research cited above suggests that the EITC program has been an important factor in increasing labor force participation. The EITC’s “work disincentive” refers to possible incremental adjustments in work effort – such as reducing weeks of work from 52 to 40, or for a couple, foregoing the earnings of one spouse for the sake of greater time devoted to work in the home.

The work disincentive does not encourage individuals to refrain from working altogether, so it does not lead to welfare dependency. An individual who has not been employed is not eligible for a dollar of EITC benefits. The adverse marginal tax rate effect in the EITC applies only to a household head or couple who are already working enough to earn at least $12,500 a year between them. Only in this range does an extra dollar incur some benefit reduction.

More work is not always better, from either a family or a social standpoint. Incremental reductions in hours could be good for the families and their children. The EITC offers families a more appealing set of choices, within the constraint that benefits are conditioned on work; in this context, government attempts to encourage greater work effort may be questioned.

Even if the marginal tax rate had a small effect on behavior, some remedy should be considered on equity grounds. It is not fair for a low-income worker to face a marginal tax rate of 50%, while one with higher income could be facing a rate of 30%.

Alleviating the marriage penalty
EITC eligibility depends on income and the number of children. For married couples, their separate incomes must be combined to determine eligibility and benefits. If a single person is eligible for benefits and gets married, the new family can be either eligible or ineligible for the credit, and its combined income can either increase or reduce their benefits.

For instance, a woman with two children and no labor income is ineligible for the credit, as is a man with wages below $30,850 and no children. But if the man and woman were to marry, then they become eligible for benefits. But for comparison’s sake, say a woman earns $20,000 and has two children – she would qualify for some EITC benefits. But if she marries anyone earning more than $10,850, then the new family is ineligible for the EITC.

Such benefit reductions or termination are usually described as a “marriage penalty.” It is of particular concern for potential EITC recipients because families with children and only one parent in the home are much more likely to be poor than are two-
parent households.

An example of a severe marriage penalty can be clearly illustrated through an examination of a pair of single-parent families with two children each. If these families each have labor earnings of $12,500, each receives the maximum EITC benefit of $3,816, so as separate families total benefits are $7,632. If the families are joined by the marriage of their household heads, however, their incomes are combined for purposes of determining the EITC benefit. The newly combined family’s benefit falls to $1,191. Half the total benefits disappear by the simple fact that EITC benefits do not increase for families with more than two children. The rest of the cut – $2,625 – results from the phase-out, since the family income of the beneficiaries has increased from $12,500 to $25,000. The total benefit reduction of $6,441 equals a decline of 84% in benefit level and 26% of the new family’s income.

One way to alleviate the penalty effect of an income increase is to widen both the plateau and phase-out ranges for the EITC. Another way to remedy this penalty would be to allow higher benefits for families with more than two children.

The overlap between solutions to the marriage penalty and alleviation of high marginal tax rates discussed previously is key to this proposal: the phase-out range must be lengthened to make some headway toward both goals. This reduces the perverse effects of rising marginal tax rates on benefits for children, preventing these benefits from falling at low ranges of income and rising at higher ones.

Instead of a penalty, a marriage bonus could be made possible under the EITC. The new family would have to have relatively low income, but it would be possible to create such a bonus by extending the phase-in range and adding benefits for additional children.

Simplifying the filing process
One barrier to using the EITC is the difficulty of understanding and applying for the credit. Any redesign should take such concerns into account and endeavor to increase, not discourage, participation in the program.

The most important task in applying for the EITC or in filing an income tax return is determining the amount of income subject to tax. Gross income, adjusted gross income, and taxable income all differ significantly under the current income tax. To complicate matters further, the EITC defines qualifying income yet another way.

To make filing easier, more consistent rules for income determination would simplify application for the EITC. If the measure of income used in determining incomes taxes is considered fair – it is supposedly designed to tax according to one’s ability to pay – then it should also be considered adequate for the EITC’s purposes.

Another important step in filing for the EITC is determining what constitutes the filing unit. Here, too, the EITC rules diverge from those of the income tax, and ways of merging the definitions would simplify filing for the credit.

A full analysis of problems resulting from the EITC’s unnecessary complexity is beyond the scope of this report, but it is clear that a simplified design that makes the credit more transparent to the taxpayer is a goal worth striving for.

A proposal for reform: the ‘Universal Unified Child Credit’
To address the inconsistencies and inequities in the EITC and other family tax benefits, this report proposes that the dependent exemption be converted into a credit and combined with the EITC, the Child Credit, and the Additional Child Credit. This new, combined credit – which could be named the Universal Unified Child Credit (UUCC) – would be available to all taxpayers with children whose income derives from employment.

Like the EITC, the UUCC would rise for an initial range of labor earnings, flatten out over an additional range, and then phase down. The difference is that it would not phase out to zero. It would settle to a minimum benefit of $1,270 per child. As noted above, this is the value of the exemption and Child Credit for taxpayers in the 28% bracket. Current eligibility restrictions based on interest and dividend income limits would remain in effect.

For six or more children, the maximum UUCC credit equals the benefits enjoyed by taxpayers in the 28% bracket who have children. For these taxpayers, no phase-out is needed. The credit amount simply remains constant as income rises beyond the end of the phase-in range.

At very high income levels the new credit could be phased out – in the same fashion as the exemption and Child Credit are under the existing system – but these phase-outs add complexity to the tax code. We prefer to maintain the minimum benefit for all high-income taxpayers. To offset the cost of this simplification, we adjust the tax brackets for those with high income (roughly the top 10% of taxpayers).

Table 3 shows the benefit levels, income milestones, and phase-down rates for this proposal. Maximum benefits are raised moderately for all taxpayers currently receiving the EITC. But under the UUCC, eligibility for the credit expands considerably to include all taxpayers. The minimum benefit level exceeds the current tax benefit of $913 per child for those in the 15% bracket. Expansion in this plan also provides new benefits to families with more than two children.

Because the phase-down is incomplete, the rate of phase-down and therefore the implied marginal tax rates are much reduced. The phase-down rates are shown in the table. As previously indicated in Table 1, the phase-out rates for families with children are 16% and 21% under the EITC.

Figure 3 is a graphic representation of this UUCC proposal. The EITC “pyramid” on the left, and the income tax benefit “steps” rising from left to right are the same as those shown in Figure 2. The dotted line represents the UUCC proposal. Benefits under the UUCC rise according to the current EITC rate but reach a higher maximum. There is a wider income range covered by the plateau, over which the maximum benefit is constant. Finally, rather than benefits phasing out, they phase down to a minimum level of $1,270 per child.
The steepness of the right side of the EITC pyramid reflects the size of the phase-out rate, and hence the marginal-tax effect. Phasing down to an amount in excess of zero and over a longer range both help to make the downslope much less steep. The phase-down ends at the benefit level for taxpayers in the 28% bracket.

Intuitively, the UUCC works by eliminating part of the EITC’s benefit reduction. The minimum UUCC benefit seamlessly merges with the tax-based benefits for children available to those in the 28% bracket. Those currently in the phase-out range would either lose benefits more slowly or not lose them at all. Those in the 15% bracket receive a tax credit not less than that enjoyed by those in the 28% bracket. If the credit is larger than the tax liability, then there would be a cash refund for the difference.

The more gradual phase-out and the new benefits for additional children reduce marriage penalties in two ways. First, a more gradual phase-out means that if two families combine their incomes, there is less benefit loss for the resulting merged families. Second, a fixed minimum credit per child, for each child, means that merged families with more than two children do not lose benefits to the same extent compared to the status quo. (See the appendix for illustrations of this proposal’s effects in terms of tax savings and alleviation of marr
iage penalties.) Marriage penalties could be further reduced by raising the maximum benefits for larger families, making the new program more expensive, or by reducing the maximums for families with one and two children.

Although not included in this report’s cost estimates, there are other possible modifications worth considering as well. One basic issue worth examining is the sort of income that should be included for determination of the credits. Under the EITC, income determination is restricted narrowly to wages and salary. Income could be extended to include at least two areas: unemployment compensation (UE) and Social Security benefits.

Unemployment compensation is financed by taxes on wages. The provision of benefits reflects the past work effort of the beneficiary. Subsidizing UE benefits is in keeping with the values the EITC was designed to promote, that is, making work pay. That unemployment should trigger a reduction of tax credits for children seems perverse.

Social Security benefits are analogous to UE benefits in terms of their reflection of past work effort. Keeping in mind the limitation on income not derived from wages, if a retiree is involved in child care – presumably caring for grandchildren – then it would seem appropriate for them to be given parity in tax benefits. According to the Census Bureau, over four million children live in households maintained by grandparents (Casper and Bryson 1998).

In the case of both UE and Social Security, taxpayers with dependents enjoy an even greater disparity of tax-based benefits if their income is sufficiently high.
In terms of other types of qualifying income, families receiving payments by state governments under the new Temporary Assistance for Needy Families program (especially when it is in the nature of labor compensation) should also be considered.

Some types of income – interest and dividends – should disqualify taxpayers from the program. Furthermore, the list of excluded income sources should be augmented to include other types of income, such as rents and royalties. At the same time, since the UUCC applies to a broad segment of the population, some raising of the limit levels for non-labor income is in order. One solution is to provide the minimum benefit to all such persons, but deny refundability on the grounds that non-labor income reflects the ownership of financial assets that should disqualify the taxpayer from receiving work-conditioned benefits.

Transforming the exemption into a credit has been proposed in the past by Steuerle (1994) and Lav and O’Hare (1998), among others. These plans all focus on improving the design of the tax code without changing the extent of progressivity or revenues. This report’s proposal has the additional goal of providing tax relief that reduces income inequality, increases the gains from work, and benefits children.

Cost of the UUCC proposal
The rough annual cost of this report’s proposal is $32 billion, which was estimated using the tax model of the Institute on Taxation and Economic Policy. (See the appendix for details of these estimates and technical notes.)

Although the focus of this report is to increase benefits for children, this proposal includes an increase in the credit for families and single persons without children. A rationale for this expansion is that fathers living apart from households where their children reside might use the credit to help finance a marriage that would reunite them with their children. More generally, such an expansion could reduce poverty by increasing labor force participation.

Perhaps the area in which this proposal is most vulnerable to criticism is its cost in terms of foregone tax revenue and explicit cash payments to individuals. Cost in this sense raises the question of fiscal policy, that is, of how the national government plans to manipulate its budget deficits and surpluses to stabilize national employment and prices.

Under current budget projections, what are called “full employment surpluses” are anticipated for the next five years (Office of Management and Budget 2000). This means that, if allowed to persist, these surpluses would push the unemployment rate above its long-run potential. Some reduction of these surpluses is appropriate.

In this context, the problem with this proposal is not that it costs too much, but that it costs too little. Projected full-employment surpluses over the next five years are estimated at $850 billion. This proposal’s cost is quite a bit less.
Naturally, other uses for the surplus will be proposed. In particular, the cost of maintaining current services in existing programs will limit the surplus available for maintaining low unemployment. The merits of all such ideas, including those in this report, ought to be compared objectively with each other as well as with the highly debatable option of using surpluses to pay off the national debt.

Merits of the UUCC proposal
As with any significant change in a program as large as the current federal income tax system, this report’s proposal has its merits and its drawbacks. The following list comprises the proposal’s benefits:

• The UUCP as proposed here lifts the incomes of a greater number of lower-income families, especially those currently below the poverty line.
• Because the UUCC would be available to most taxpayers below median income levels, it would reduce after-tax income inequality.
• The proposal radically reduces the implicit marginal tax rates faced by many low-wage workers.
• The proposal increases parity between tax-based benefits for children available to taxpayers at different income levels, and, more generally, it improves the sensitivity of the tax code to differences in the ability to pay taxes stemming from family size.
• It alleviates the marriage penalty inherent in the EITC, since the phase-out rate is reduced and the range of phase-out is wider.
• It provides some parity for larger families by creating new benefits for taxpayers with more than two children.
• It mitigates the regressivity of Social Security and Medicare financing (though both programs are progressive when benefits are taken into account), and it speaks to the concerns of younger workers whose payroll taxes will be financing a disproportionately large retirement population.
• It simplifies filing for the EITC by merging the credit directly with the income tax structure, and makes the system as a whole more transparent and comprehensible to taxpayers.
• It remedies the neglect of children whose parents are unemployed. It also benefits grandparents maintaining households with children who depend on Social Security benefits.
• It increases the value of the Dependent Care Credit, which is currently non-refundable. The replacement of dependent exemptions with the UUCP reduces the taxpayer’s “zero bracket level” and increases positive tax liability before the UUCP is subtracted. This creates a type of double benefit, first because the UUCP exceeds the present value of the exemption and Child Credit, and second because of the increased usability of the Dependent Care Credit. This would follow for other non-refundable credits as well, such as those for education, adoption, and care of the elderly and disabled.
• It encourages workers to report their income for the sake of recei
  ving the benefits of the credit, thereby contracting the “underground economy.” Such income reporting facilitates government efforts to enforce loan commitments and child support obligations that apply to the taxpayer. It also strengthens the Social Security and Medicare programs by increasing contributions to the trust funds for those programs.

Possible drawbacks of the UUCC proposal
One potential drawback of the proposal pertains to the reduction of the phase-out rate. For a given maximum benefit level, reducing the phase-out rate necessitates enlarging the phase-out range. A new group of taxpayers presently ineligible for the program will get reduced tax liability but an increased implicit marginal tax rate. It is possible that such a taxpayer would reduce his or her labor earnings, though the low phase-out rates under the UUCC should help mitigate this.

Some might view an incremental reduction in work with disfavor. On the other hand, one could imagine situations where such a reduction was more favorably regarded. A pair of parents who each have full-time jobs could decide to reduce their work time to better meet child-rearing responsibilities.
A criticism of the EITC voiced by conservatives is that it is too high a subsidy – it compensates some persons by much more than their tax liability. One response is that the credit defrays payroll taxes, not just individual income taxes.

Another response to the conservative criticism is that equality of opportunity, as opposed to “equality of result,” is a conservative value. Children’s opportunities in life clearly depend on family income. If personal responsibility is not in question, then family income depends largely on the fate of parents in the labor market. Opportunity in this context clearly varies to an important extent with factors beyond the worker’s control. A work-based refundable tax credit mitigates against such factors, enhancing opportunities without raising issues of moral hazard.

The criticism most often voiced by those on the left is that the EITC subsidizes low-wage employers. The magnitude of the effect is indeed worthy of consideration. Research shows that the credit raises income and reduces poverty. This means that the credit supplements wages, rather than allowing employers to pay less for the same work. In actuality, substituting the credit for wages is not an easy task for an employer.

Low-wage workers are typically paid by the hour, and their employment is often irregular. EITC benefits are provided on an annual basis and vary according to family characteristics. Ninety nine percent of EITC beneficiaries receive the credit in one lump sum after filing their income taxes. The employer has no way of knowing who is eligible for the credit, nor how much they receive. An employer could make educated guesses, but imputing a worker’s EITC benefit and translating that into an hourly wage reduction would not be a simple task.
Insofar as the labor market is competitive, an employer who undercuts the going wage could lose workers to firms willing to pay a bit more, up to the value-added these workers bring to the firm. So there is a plausible countervailing pressure against the substitution of the credit for wages.

Another criticism of the program is that it pushes more people into the low-wage labor market, intensifying competition for jobs and putting downward pressure on wages for that sector. Two programs – Aid to Families with Dependent Children and General Assistance – have traditionally provided a safety net in the low-wage sector, but the welfare reform process is radically transforming these programs, making abstention from the labor market much more difficult. More importantly, general public disapproval of failure to maintain employment on the part of able-bodied adults has been the main factor giving rise to welfare reform and the progress of the EITC. Insofar as this disapproval remains a political constraint on reform, work-contingent benefits like the EITC are the only way to increase cash income for families who depend on jobs paying low wages.

Some of the specifics of our plan are also open to criticism. One might prefer greater increases in benefits. But to some extent, what is given must also be taken away as income rises, down to some minimum benefit level. The higher the maximum, the steeper the phase-out and the greater the potential for work disincentives and marriage penalties. A more gradual phase-out for a given benefit raises the cost of the plan. As things stand, the cost of our proposal is considerable for a single-tax reform measure.

The structure of our proposal makes benefit increases easy to implement. One merely lengthens the phase-in ranges in tandem. This increases the maximum benefit levels. Families of all sizes are already accommodated. The phase-out ranges could be expanded, increasing the cost of the reform, or contracted somewhat to reverse effect.

Alternatively, one might appreciate the simplicity of a credit that rose to a set level and remained there for all taxpayers with sufficient labor income, as our own plan does for families with six or more children. Here the difficulty is that budget considerations could lead to a small universal credit that would result in benefit reductions or tax increases for many families.

Conclusion
In its fiscal year 2001 budget, the Clinton Administration proposed a lengthy list of tax cuts, totalling approximately $350 billion (Congressional Budget Office 2000). It is not likely to be approved by a Republican Congress, nor to be the focus of debate in this election year. By contrast, a single proposal, like the Universal Unified Child Credit, devoted to a profound need, might have a better chance. This plan’s cost is in the same range as the administration’s tax cuts, and is far less expensive than the tax plan proposed by Governor George W. Bush.

Although the concept of work-conditioned tax credits for children enjoys bipartisan political support, the EITC has always been subject to more criticism and scrutiny than any other part of the individual income tax code. To some extent, EITC beneficiaries are politically isolated and stigmatized. Universalizing the credit could foster a more balanced consideration of its merits.
The goal of this report is to stimulate a constructive debate on how to improve tax equity for families with children. The proposed Universal Unified Child Credit makes work pay more for those with the greatest need, which should be the priority for tax reform and relief.

April 2000

The authors would like to thank Randy Albelda, Eileen Appelbaum, Joel Blau, Chauna Brocht, Michael Ettlinger, John O’Hare, Waltraud Schelkle, John-Karl Scholz, and Eugene Steuerle for their helpful comments.

Appendix

Estimates discussed in this section were prepared by Michael Ettlinger of the Institute on Taxation and Economic Policy (ITEP). ITEP maintains a large-scale, highly regarded tax simulation model.

Cost of the UUCC
The use of the term “cost” is routine in tax debates, but it has a double meaning that needs to be borne in mind. A tax cut “costs” the federal government in terms of revenue, but the lack of such a tax cut costs the prospective beneficiary of the policy. Accordingly, our UUCC costs the federal government, but the lack of a UUCC reform will cost low-income families with children.

For fiscal year 1999, the Earned Income Tax Credit cost the federal government $30.4 billion (Office of Management and Budget 2000). Of this, $4.8 billion3 was defrayed income tax
liability, and $25.6 billion took the form of cash refunds in excess of income tax liability.

We estimate the total increase in cost of the UUCC to the federal budget at $32 billion for tax year 1999.4 This is a net cost, because under our plan some persons’ tax liabilities increase. The gross tax cut is $35 billion, combined with a $3 billion tax increase.

Of the $32 billion net tax cut, $13.5 billion consists of literal tax refunds and the remainder – $18.5 billion – is a net reduction in income taxes. As noted above, $4.8 billion of the current EITC is a literal income tax cut. In addition, the existing federal child credit defrays $19.4 billion of tax liability in 1999 (OMB 2000). There are also tax savings from the dependent exemption. In 1996, exemptions for all taxpayers totaled nearly $600 billion (IRS 1999). Valued at the minimum marginal tax rate of 15%, these would have provided tax savings of roughly $90 billion. Against this background, as noted above, the net increase in income tax reduction resulting from the UUCC is $18.5 billion.

Because the EITC is refundable, it is said to be excessive or wrongly targeted as a device for cutting taxes. But if one considers the payroll tax, then part of the cash refund could be interpreted as payroll tax relief. Under present law, $3.3 billion of the EITC offsets payroll tax liability. The increase in such relief under the UUCC is $3.8 billion, hence the increase in refunds net of both the payroll and income tax is $14.7 billion (18.5 minus 3.8).

Our plan includes an increase in benefits for taxpayers with no children by cutting their taxes an additional $5.2 billion. As noted in the report, such a cut would be expected to increase labor force participation and reduce poverty. But it is not central to the issue of benefits for children, so budgetary considerations could justify its exclusion from the reform. That change would reduce the total cost of the reform from $32 billion to about $27 billion.

The replacement of the dependent exemption with a refundable credit has some unexpected effects. As noted in the text, since tax liability before applying credits increases, the non-refundable credits for dependent care, elderly care, education, and foreign taxes become more valuable. The largest such change of this type is for the dependent care credit; the ITEP model estimates it at $75 million annually. This change is included in the overall cost figures.

A small portion of the $3 billion in tax increases under the UUCC is created by anomalies affecting taxpayers with low income; these would be remedied with more detailed rules than we have specified in the paper. An example is a higher tax liability for some low-income taxpayers with pension distribution income: the elimination of the dependent exemption increases their tax liability more than the UUCC diminishes it. These taxpayers are in the “phase-in” range, but the extent of the credit phased in is less than the value of the exemption in offsetting non-labor income.

The main source of increase stems from the elimination of existing phase-outs of the dependent exemption and child credit. This elimination benefits high-income taxpayers with children. We offset the cost of this tax cut by changing the brackets associated with the 36% and 39.6% rates. Some high-income taxpayers, particularly those without children, end up with higher tax liability under the UUCC plan.

Since the tax base grows over time, the annual cost of the UUCC rises as well. From 1999 to 2000, the estimated growth in cost is 4.3%. Since recent economic growth has been unusually rapid, a conservative five-year projection suggests an aggregate cost of $174 billion, roughly equivalent to the budget surplus for a single year. The 10-year total is approximately $390 billion.

Distributional effects of the UUCC
Table A-1 shows average tax cuts and tax increases under the UUCC. Taxpayers are divided into quintiles, starting with the lowest. The third column shows the change in tax liability (defined here as the combination of tax cut and cash refund) as a percent of income. These percentages are small, showing that it is not possible to redistribute much income with only $32 billion. The “average change” column shows the average of tax cuts and tax increases; note that it averages tax cuts and tax increases within each income group, so the resulting figure is not very revealing. More relevant is the average tax cut for those receiving a tax cut and the average increase in tax liability for those receiving an increase. The latter is an exceedingly small number for the bottom 80%, since virtually none of these taxpayers are “losers” under this plan.

Most of those receiving tax cuts have incomes between $13,600 and $64,900, or the 21st through 80th percentiles. The high percentages of “winners” in the top 10% are the result of our elimination of the phase-outs of the existing child credit and dependent exemption. By the same token, tax hikes in the top decile are the result of our change in the tax brackets. The net effect for high-income taxpayers is to raise taxes by $3 billion, which offsets a small part of the tax cut for those with lower incomes. The change in effective tax rates for those in the top decile is a 10th of a percent or less.

Table A-2 shows the same results but with different income ranges, so as to allow a better pinpointing of the pattern of tax cuts by income level. “Winners” in the $40,000-50,000 income range receive the biggest cut.

The benefits of the cut could be shifted down by raising the maximum benefits and allowing a phasing-down to the minimum benefits more rapidly. As noted in the report, this change would reduce the improvement in marginal tax effects and reduction of marriage penalties in our plan.

Table A-3 shows the proposed changes in brackets for high-income taxpayers. Note that these changes are defined in terms of taxable income. Taxable income is the remainder after subtracting standard or itemized deductions and exemptions, among other items, from adjusted gross income (AGI). Thus, for a married couple filing jointly, the 36% bracket begins at $123,699, but since its standard deduction is $7,200, this couple’s AGI must be at least $130,899. In practice, most such couples will have itemized deductions that far exceed $7,200, so their AGI would be that much higher. Furthermore, AGI entails subtractions from gross income for such items as contributions to individual retirement accounts. The upshot is that the taxable income numbers in this table reflect much higher gross income levels.

Effects on increases in labor earnings

One of the prime motivations for this reform is to reduce losses to total income resulting from an increase in labor income. Figure A-1 illustrates some changes in disposable income if gross labor earnings rise from $10,000 to $15,000 under different scenarios. We assume for these calculations that child care costs 10% of gross income with one child and 15% with two or more. We include the effect of the payroll tax (both employer and employee), and assume job-related expenses of 5%. Although child care costs are assumed in any case, we alter the assumption that the taxpayer takes the dependent care credit. (If the child care provider is unlicensed, the taxpayer cannot use the credit.)

The first bar, which represents current law, shows that the $5,000 income increase would net the taxpayer between about $2,500 and $3,000, depending on the number of children. Under the UUCC, the amounts are generally m
uch higher, from $3,464 for a household with two children to $4,867 to a household with four. These increases are smaller if the dependent care credit is not available to the household. For a household with just one child, the UUCC actually reduces the increase in disposable income from $3,092 to $2,736, due to the anomalies in the plan discussed above. Further refinement of the plan could raise that amount.

Effects on marriage penalties
Table A-4 shows some effects of the plan on the so-called marriage penalties. The first column describes different situations, in terms of income and number of children, for a single household head. The subsequent columns show reductions in the marriage penalty under this plan for that person and his or her spouse at different income levels.

For example, the first case is a head of household with income of $8,000 and one child. If he or she marries and ends up in a family with income of $20,000, the reduction in the marriage penalty, relative to current law, is $413 annually. If family income is $25,000, the penalty reduction is $866. The last scenario in the first column is a household head with $15,000 of income. With two children, marrying “into” a family that will have a total income of $25,000 means a penalty reduction of $589.

This plan does not eliminate marriage penalties altogether. This elimination is impossible if there is any phase-down of benefits, or if benefit levels are not strictly proportional to the number of children. Either of these measures would make the cost of the plan prohibitive, unless the credit were reduced below current levels. The intention of this plan is to provide tax relief, which means respecting maximum benefit levels under the EITC and current-law values for the dependent exemption and child credit. A lower uniform credit would amount to a tax increase for those with the lowest labor incomes.

Conclusion 
Our account of the effects of the UUCC on earnings increases and marriage penalties is only meant to be suggestive, not conclusive. There is plenty of room for more exhaustive analysis. We also welcome additional efforts to model this type of reform and refine our cost estimates. This is not a simple task. In general, we hope interest in the UUCC stimulates more research on tax reform, “from the bottom up.”

Endnotes

1.Since 1981, exemptions have been adjusted for inflation. Even so, they are worth much less than in prior decades. Steuerle (1983) notes that in 1984, the inflation-adjusted value of the exemption in 1948 would have grown to $5,600. Today, a 1948 exemption would be worth nearly $9,000. This is important because the treatment of exemptions and child credits illustrates that the current tax code is not sensitive to how differences in family size affect ability to pay taxes (Steuerle 1994).

2. In the credit’s “phase-in” range, it is like a “negative tax,” that is it supplements wages. While phasing in, an extra $100 of wages brings an extra $40 in benefits for a family with two children.

3. By convention, an EITC offset to the self-employment tax is counted as defrayed tax liability and scored as a tax cut in budget presentations. So a small part of the $4.8 billion does not defray individual income tax liability per se.

4. The tax year is a calendar year, while the fiscal year runs from the last quarter of one year through the first three quarters of the succeeding year. Accordingly, our commingling of numbers from these two different time periods is an approximation of the actual magnitudes in question.

References
Barthold, Thomas A., Thomas Koerner, and John F. Navratil. 1998. Effective Marginal Tax Rates Under the Federal Individual Income Tax: Death by a Thousand Pricks? National Tax Journal. Vol. 51, No. 3, pp. 553-64.

Casper, Lynn M., and Kenneth R. Bryson. 1998. “Co-resident Grandparents and Their Grandchildren: Grandparent Maintained Families.” U.S. Census Bureau, Population Division, Working Paper No. 26. Washington, D.C.

Congressional Budget Office. 2000. An Analysis of the President’s Budgetary Proposals for Fiscal Year 2001: A Preliminary Report. Washington, D.C.

Dickert, Stacy, Scott Houser, and John Karl Scholz. 1995. The Earned Income Tax Credit and Transfer Programs: A Study of Labor Market and Program Participation. Tax Policy and the Economy. No. 9. Cambridge, Mass.: M.I.T. Press.

Eissa, Nada, and Jeffrey Liebman. 1996. Labor Supply Response and the Earned Income Tax Credit. Quarterly Journal of Economics. Vol. 111, No. 2.

Ellwood, David T. 1999. “The Plight of the Working Poor.” Paper presented at The Brookings Institution, Children’s Roundtable, No. 2, Washington, D.C., November.

Giannarelli, Linda, and C. Eugene Steuerle. 1995. “The True Tax Rates Faced by Welfare Recipients.” Paper presented at National Tax Association Proceedings, San Diego, Calif., 1995. Greenstein, Robert. 1998. The Earned Income Tax Credit and Error Rates. Washington, D.C.: Center on Budget and Policy Priorities.

Greenstein, Robert. 2000. Should EITC Benefits Be Enlarged for Families with Three or More Children? Washington, D.C.: Center on Budget and Policy Priorities.

Greenstein, Robert, and Isaac Shapiro. 1998. New Research Findings on the Effects of the Earned Income Tax Credit. #98-022. Washington, D.C.: Center on Budget and Policy Priorities.

Holtzblatt, Janet and Robert Rebelein. 1999. Measuring the Effect of the EITC on Marriage Penalties and Bonuses. Washington, D.C.: Joint Center for Poverty Research.

Internal Revenue Service. 1998. Statistics of Income—1986, Individual Income Tax Returns. Washington, D.C.: Internal Revenue Service.

Lav, Iris J. 1999. Alleviating Marriage Penalties in the EITC. Washington, D.C.: Center on Budget and Policy Priorities.

Lav, Iris J., and John F. O’Hare. 1998. “Cleaning the Code: Strategies for Simplifying the Individual Income Tax.” Paper presented at National Tax Association proceedings, November.

Liebman, Jeffrey B. 1999. The Optimal Phaseout Tax Rate for the Earned Income Tax Credit. Washington, D.C.: Joint Center for Poverty Research.

McCubbin, Janet. 1999. Non-Compliance and the Optimal Design of the Earned Income Tax Credit. Washington, D.C.: Joint Center for Poverty Research.

Meyer, Bruce, and Dan T. Rosenbaum. 1999. “Welfare, the Earned Income Tax Credit, and the Labor Supply of Single Mothers.” National Bureau of Economic Research, Working Paper No. 7363. Cambridge, Mass.: NBER.

Office of Management and Budget. 2000. Federal Budget, Fiscal Year 2001. Washington, D.C.: U.S. Government Printing Office.

Scholz, John-Karl. 1994. The Earned Income Tax Credit: Participation, Compliance, and Antipoverty Effectiveness. National Tax Journal. Vol. 47, No. 1, pp. 59-81.

Scholz, John Karl. 1997. Testimony before House Committee on Ways and Means, U.S. House of Representatives, May 8.

Smeeding, Timothy M., Katherine E. Ross, Michael O’Connor, and Michael Simon. 1999. The Economic Impact of the Earned Income Tax Credit (EITC): Consumption, Savings, and Debt. Washington, D.C.: Joint Center for Poverty Research.

Steuerle, Eugene. 1983. “The Tax Treatment of Households of Different Size.” In Rudolph G. Penner, ed., Taxing the Family. Washington, D.C.: American Enterprise Institute.

Steuerle, Eugene. 1994. Why I Favor Child Credits or Allowances. Tax Analysts.

U.S. Census Bureau. 2000a. Low Income Uninsured Children by State. Washington, D.C., U.S. Government Printing Office.

U.S. Census Bureau. 2000b. Poverty 1998. Washington, D.C.: U.S. Government Printing Office.