urban institute nonprofit social and economic policy research

Noncustodial Parents, Child Support, and the Earned Income Tax Credit

Read complete document: PDF


PrintPrint this page
Share:
Share on Facebook Share on Twitter Share on LinkedIn Share on Digg Share on Reddit
| Email this pageE-mail
Document date: November 28, 1997
Released online: November 28, 1997

Support for this research comes from the Charles Stewart Mott Foundation. The opinions expressed in this paper are those of the authors and do not necessarily reflect the views of the Urban Institute or its funders. The authors wish to thank Leonard Burman, Robert Lerman, Eugene Steuerle, and Leslie Whittington for their comments regarding this work. Thanks are also extended to Nikki Blasberg who conducted the programming and research assistance for this paper.

Note: Those wishing to print this document should use the PDF version.


Contents

Abstract

Introduction

Section 1: Taxes, Child Support, and Financial Well-Being

Tax Relief for Families with Children
Child Support Guidelines
Effect of Guidelines and Taxes on Low Income Families
Who Should Bear the Cost of Marital Dissolution and Non-Marital Birth?

Section II: The EITC

Legislative History of the EITC
Eligibility Criteria, Benefit Levels, and Work Incentives and Disincentives
Characteristics of EITC Eligibles
Characteristics of Eligible Custodial Mothers and Noncustodial Fathers

Section III: Policy Options and Financing Methods

Three Options for Expanding the EITC
Costs and Benefits Under Each Policy Option
Financing the Expansion
Combined Effects of Policy and Financing Options on Relative Well-Being

Section IV: Behavioral Effects

Incentives for Payment of Child Support
Incentives for Receipt of Child Support
Incentives Affecting Work: Noncustodial Parents
Incentives Affecting Work: Custodial Parents
Incentives Affecting Marriage
Compliance with EITC Rules

Summary and Conclusions

References

Appendix


List of Tables

Table 1: The Effect of Federal Income Taxes, EITC, and Child Support on the Noncustodial Parent, Custodial Family, and Married Parent Family Assuming $12,000 in Earnings for Each Parent, and One Child
Table 2: Poverty Level of Noncustodial Parent, Custodial Family, and Married Parent Family, Before and After Taxes, EITC, and Child Support, Assuming $12,000 in Earnings for Each Parent, and One Child
Table 3: Poverty Level of Noncustodial Parent, Custodial Family, and Married Parent Family, Before and After Taxes, EITC, and Child Support, Assuming $12,000 in Earnings for Each Parent, and Two Children
Table 4: Poverty Level of Noncustodial Parent, Custodial Family, and Married Parent Family, After Federal Income and Payroll Taxes and Child Support, For Different Combinations of Noncustodial and Custodial Parent Earnings, and One Child
Table 5: Poverty Level of Noncustodial Parent, Custodial Family, and Married Parent Family, After Federal Income and Payroll Taxes and Child Support, For Different Combinations of Noncustodial and Custodial Parent Earnings, and Two Children
Table 6: 1996 Earned Income Credit Parameters
Table 7: Characteristics of 1996 EITC Eligibles
Table 8: Characteristics of EITC Eligible Custodial Mothers and Noncustodial Fathers
Table 9: Costs and Benefits of Increasing the EITC to Low-Income Noncustodial Fathers Who Pay Child Support
Table 10: Poverty Level of Noncustodial Parent, Custodial Family, and Married Parent Family After Taxes, EITC, and Child Support, Under Different EITC Scenarios, Assuming $12,000 in Earnings for Each Parent

Abstract

As welfare reform moves families from reliance on welfare to reliance on work, child support and the Earned Income Tax Credit (EITC) play an increasingly important role in the financial well-being of low-income single parent families with children. The first section of this paper examines the effect of child support, the EITC, and the tax system on the relative well-being of low-income noncustodial parents and custodial families. The authors find that a low-income noncustodial parent must have earnings 50 to 100 percent higher than the custodial parent in order to pay child support and taxes and enjoy the same standard of living as the custodial family. Noncustodial parents receive little of the tax relief available to families with children, and typical state child support guidelines do little to redress this. The authors explore three options for expanding the EITC to greater assist low-income noncustodial parents who pay child support, and discuss two options for partially or fully funding this expansion.


Introduction

As welfare reform moves families from reliance on welfare to reliance on work, two programs play an increasingly important role in the financial well-being of low income single parent families with children—the federal child support enforcement system (CSE), and the earned income tax credit (EITC). Each program was created in the 1970s and has been expanded considerably since that time. The mission of CSE is to ensure that children eligible for child support receive it. The federal office of child support enforcement (OCSE) oversees state child support plans, and funds 66% of state child support enforcement costs. In 1995, federal administrative expenditures on child support enforcement were $2 billion, and the system collected $10.8 billion in child support payments on behalf of more than 3.4 million families (Green Book 1996). The purpose of the EITC is to provide tax relief and subsidize the incomes of low income working families, thus providing an incentive for work. In 1995, 20 million taxpayers received a total of $26 billion in EITC benefits (Internal Revenue Service 1997). Originally available only to families with children, low income taxpayers without children became eligible for a small credit in 1995.

In this paper, we consider the effect of child support and federal income taxes on the financial well-being of low-income noncustodial parents and custodial families, and explore three options for expanding the EITC to provide tax relief to low income noncustodial parents who pay child support. In Section 1, we describe the tax treatment of low income noncustodial and custodial parents, discuss state guidelines used in setting child support awards, and explore the combined effect of taxes and child support on the financial well-being of low-income custodial and noncustodial parents. We find that when low income custodial and noncustodial parents have similar levels of earnings and the noncustodial parent pays child support set according to typical state guidelines, the combination of child support payments and federal income taxes yields a lower standard of living for the noncustodial parent than for the custodial family. Only when the noncustodial parent's earnings exceed the custodial parent's by 50 to 100 percent, do low-income noncustodial parents who pay child support fare as well as the custodial families they help to support.

There are two reasons that low-income noncustodial payers fare worse than recipients with similar levels of earnings: tax benefits directed to families with children go almost entirely to custodial parents; and state child support guidelines do little to redress this. Given our country's greater willingness to assist poor children than to assist able-bodied adults, it is not inconsistent that low income noncustodial parents be asked to bear the brunt of the cost of marital dissolution and non-marital birth. Although it would be most equitable for the custodial and noncustodial parent to share these costs equally, this would result in a lower standard of living for the children. Nevertheless, the fact remains that child support and taxes swallow a substantial chunk of the income of low income noncustodial parents who pay child support. Despite helping to support their children, most noncustodial parents receive little tax relief for doing so.

We propose three options for expanding the EITC to provide tax relief to noncustodial parents who pay child support. We choose the EITC as the vehicle for tax relief for two reasons: it is targeted to low income noncustodial parents; and it provides an incentive to work and pay child support. In Section II, we discuss the legislative history of the EITC, EITC eligibility and benefit rules, and characteristics of taxpayers currently eligible for the EITC. In section III, we present three options for expanding the EITC and two methods for (partially or fully) financing the expansion. Potential behavioral consequences of the policy and financing methods are discussed in section IV. We present our summary and conclusions in section V. An appendix describes the data used in this paper (the Survey of Income and Program Participation), the method used to determine EITC eligibility from these data, and a validation exercise to assess data reliability.


Section 1: Taxes, Child Support, and Financial Well-Being

In this section, we discuss the combined effect of taxes and child support on the financial well-being of low-income noncustodial parents and custodial families. We begin by describing the primary means by which the federal income tax system directs tax relief to families with children. Next, we discuss the three major types of state guidelines used in setting child support awards, and consider the extent to which these guidelines take into account federal income taxes. We then explore the effect of federal income taxes and child support payments on the financial well-being of hypothetical low-income noncustodial parents and custodial families. We conclude this section by considering the issue of the optimal distribution of income between the noncustodial parent and the custodial family.

Tax Relief for Families with Children

Federal income tax law treats families with children more favorably than similar income families without children. The three primary features of the tax law geared toward providing tax relief to families with children are the personal exemption, head of household filing status, and EITC. The $500 per child credit, passed under the Tax Payer Relief Act of 1997, will provide additional tax relief to all but the lowest income families with children. For the most part, noncustodial parents do not benefit from these provisions. We discuss each of these provisions in greater detail below.

In 1996, the personal exemption reduced taxable income by $2,550 (phased out at high income levels) for each dependent child. In general, the custodial parent is entitled to the dependent exemption if the custodial and noncustodial parents together paid over half the child's support. The noncustodial parent is entitled to claim the child as a dependent if the above condition is met and: 1) the custodial parent relinquishes the exemption by signing federal income tax form 8332; 2) the noncustodial parent was granted the exemption under a post-1984 divorce decree or separation agreement; or 3) a pre-1984 divorce decree grants the noncustodial parent the exemption, and the noncustodial parent paid at least $600 for the child's support in 1996. Relatively few noncustodial parents—811 thousand in 1993—claim the dependent exemption for a child living with a former spouse. Furthermore, most noncustodial parents claiming the dependent exemption are in the middle or upper income ranges. Fewer than 12 percent have adjusted gross income below $15,000.1 Finally, many of the dependency claims of noncustodial parents are incorrect. In analyzing a sample of noncustodial parents who failed to pay child support in 1993, the Inspector General of the Department of Health and Human Services (1996) found that 25% incorrectly claimed the dependent exemption when filing their income taxes.

Head of household filing status is available to any unmarried custodial parent who provides over half the cost of keeping up her home, regardless of whether she claims her children as dependents. The noncustodial parent cannot file as head of household, even if he is the primary supporter of his children. By filing as head of household, the custodial parent receives a higher standard deduction—$5,900 in 1996—than the $4,000 she would have received as a single filer. In addition, middle and upper income individuals filing as head of household benefit from a more favorable rate schedule, and high income individuals benefit from a higher threshold for the phase-out of the personal exemption.2

The EITC is a refundable tax credit available to low income families and individuals with earned income. Until 1995, the EITC was available only to families with a "qualifying child." In order for a child to qualify a parent for the EITC, the child must have resided with the parent for more than half the year. Thus, a child can qualify a custodial parent for the EITC even if the noncustodial parent claims the child as a dependent. The child cannot qualify the noncustodial parent for EITC, since the child does not reside with the noncustodial parent for more than half the year. In 1995, EITC eligibility was extended to individuals without qualifying children. However, the maximum credit is quite small ($323 in 1996) relative to the maximum credit for families with qualifying children ($2,152 for one qualifying child, and $3,556 for two or more qualifying children). Furthermore, the EITC is completely phased out by the time earnings reach $9,500 for individuals without a qualifying child, but is not phased out until $25,078 for families with one qualifying child, and $28,495 for families with two or more qualifying children.

Finally, the $500 per child credit, enacted under the Taxpayer Relief Act of 1997, benefits the parent who takes the child as a dependent—in most cases, the custodial parent. The child credit does not benefit the lowest income families—those with zero taxable income after deductions and exemptions. For example, had the $500 per child credit been in effect in 1996, a single parent with one child would have required adjusted gross income of more than $11,000 to benefit even partially from the credit. To receive the full credit, she would have required adjusted gross income of about $14,300. We do not include the child credit in results presented in this paper.

Child Support Guidelines

Since 1989, the federal government has required that states have mandatory judicial guidelines for setting child support awards. A guideline is a formula that calculates the child support award as a function of the noncustodial parent's income, or the combined incomes of the custodial and noncustodial parent. Each state chooses its own child support guideline. Judges and other officials who set child support awards are allowed to deviate from the guidelines, but are required to record the reason for the deviation.

There are three major types of child support guideline—income shares, percentage of income, and Melson. In 1990, 33 states or territories were using the income shares guideline; 17 were using the percentage of income guideline; and 3 were using the Melson formula (Bassi and Barnow 1993). Although states can be broadly classified into these three guideline types, within each type there is considerable variation in details such as income definition, award levels, and treatment of extremely high and low income parents.

The income shares guideline is based on the premise that children are entitled to the same share of each parent's income as they would have received had the parents resided together.3 A total award is calculated based on the combined income of the two parents. The award is then allocated between the parents in proportion to their incomes. The noncustodial parent must pay his or her portion through child support—the custodial parent's portion is assumed to be fulfilled through caring for the child. An equivalent way of expressing the income shares guideline is that the noncustodial parent's obligation is equal to a percent of his or her income, in which the percent varies with the combined incomes of the two parents. The variation in the percent is intended to reflect typical changes in the percent of income spent on children in in-tact families as income rises.

The percentage of income guideline is simpler to administer than the income shares guideline because it does not take into account the income of the custodial parent. The noncustodial parent's child support obligation is set as a percentage of his or her income. In some of the percentage of income states, the percent varies with the level of the noncustodial parent's income—in others, it remains constant at all income levels.

Finally, the Melson formula sets the child support award at a fixed percent of income beyond a minimum "self support reserve." As with the income shares formula, the percent is determined by the combined income of the noncustodial and custodial parent.

Taxes may be taken into consideration in setting child support guidelines either implicitly or explicitly. In about two thirds of the income shares states and over a third of the percentage of income states, pre-tax income is used in determining the child support award. These states, to the extent that they account for the impact of taxes, do so implicitly—through adjusting the amount of the award for each income level. The adjustments are based on assumptions about the relative incomes of the parents and about who claims the child as a dependent (usually, the custodial parent). We suspect that the impact of the EITC on the relative incomes of noncustodial parents and custodial families was ignored in the development of child support guidelines—possibly because the credit was much smaller at the time most guidelines were being developed (prior to the 1990 expansions).

For the remaining income shares and percent of income states, and for all of the states using Melson guidelines, taxes are taken into account explicitly— after-tax income is used in calculating the child support award. Because after-tax income is used, the award is typically set as a higher percent of income than is necessary for pre-tax income.

Effect of Guidelines and Taxes on Low Income Families

A number of studies have shown that divorce typically causes a drop in the standard of living of mothers and children and an increase in the standard of living of fathers (Weiss 1984, Duncan and Hoffman 1985, Weitzman 1985, Stirling 1989, Smock 1994, and Sorenson 1992).4 In this paper, however, we show that low income noncustodial parents experience a lower standard of living, after paying child support and taxes, than do custodial families with similar earnings. Since our finding appears to contradict findings from these earlier studies, a few words of explanation are in order before proceeding.

First, our findings are not directly comparable with findings from these studies since we are focusing on a special group—low income noncustodial parents—under a special circumstance—full payment of a child support award set under typical state guidelines. Studies of the economic consequences of divorce have focused on the average actualoutcomes of custodial families and noncustodial parents. Thus, they incorporate the effects of low and nonexistent child support payments and the experiences of middle and upper income divorcing couples.5 Furthermore, the data used in these studies are old—four of the studies rely on data from the 1970s, and two rely on data from the early to mid 1980s. Thus, they do not capture more recent changes in the child support system, including the implementation of presumptive state guidelines. Also, the studies do not consider the effect of taxes or the EITC on the financial situation of the two parents. Finally, the studies do not investigate the relative well-being of fathers and mothers in cases of non-marital birth.6 Since these cases now account for about one third of all custodial mothers and one half of poor custodial mothers (Bureau of the Census 1991), results from studies based on divorcing couples are ill-suited for generalizing about the entire single parent population.

Studies of the economic consequences of divorce have drawn attention to problems affecting many custodial parents and have spurred reforms in the child support enforcement system. Unfortunately, many custodial mothers continue to remain poor, and many noncustodial fathers still do not pay child support. Our focus in this paper, however, is on what happens in the case of low income parents who "play by the rules" by working and paying child support. Although these parents may comprise only a small part of the low-income noncustodial parent population, they are of the upmost importance. If outcomes for these parents are unreasonable, then it will be difficult to encourage others to play by the rules.

As illustrated below, we find that if each parent works at a low-wage job earning between $9,000 and $18,000, the standard of living of both the noncustodial parent and the custodial family falls after divorce (or in the absence of marriage). However, the standard of living of the noncustodial parent falls the most. Child support and the EITC raise the standard of living of the custodial family relative to the noncustodial parent. In some cases the combination of child support and EITC makes the custodial family better off than prior to the divorce.

Table 1 illustrates the effect of taxes and child support on a noncustodial parent and custodial family in which each parent earns $12,000 and there is one child. The table is based on 1996 federal tax law and poverty thresholds. The child support award falls approximately in the middle of the range of awards calculated under the guidelines of income shares states that use pre-tax income in the guideline calculation. For comparison, we show taxes and the standard of living of the family if the parents were married.

Because the earnings of the two parents are equal, each pays $918 in payroll taxes (Social Security and Medicare taxes). However, since the custodial parent receives the higher head of household standard deduction and claims the child as a dependent, federal income taxes are much higher for the noncustodial parent ($818) than for the custodial parent ($150). In addition, the child makes the custodial parent eligible for $2,096 in EITC benefits. Since the noncustodial parent's earnings exceed the maximum for eligibility for the childless EITC, he receives no EITC. Had the parents been married, they would have received a small EITC benefit ($173) since their combined incomes place them far into the phase out range of the EITC.

After taxes and the EITC, the income of the noncustodial parent has fallen to $10,265, whereas the income of the custodial family has increased to $13,028. If the noncustodial parent pays the full $2,001 in annual child support, he has only $8,264 after taxes and child support. The custodial family, on the other hand, has $15,029.

The bottom panel of table 1 shows 1996 poverty thresholds and the percent of poverty, under different income definitions, for the noncustodial parent, custodial family, and married parent family. If we define income as income prior to taxes, EITC, and child support, the custodial family fares worse than the noncustodial parent. Married, the family would have had income equal to 192 percent of poverty. Separate, the incomes of the noncustodial parent and custodial family fall to 147 and 114 percent of poverty respectively. Taxes and the EITC virtually equalize the standard of living of the noncustodial parent (126 percent of poverty) and custodial family (123 percent of poverty). From this level, child support lowers the standard of living of the noncustodial father to 101% of poverty, while raising the standard of living of the custodial family to 142% of poverty. The combined effect of federal income taxes, the EITC, and child support, is to roughly reverse the standards of living of the noncustodial parent and the custodial family. Had there been two children instead of one, taxes, the EITC, and child support would have actually made the custodial family better off than the married parent family.

The impact of taxes, the EITC, and child support on the standard of living of the noncustodial parent and custodial family in the above example are illustrated graphically in table 2. The table consists of six charts showing the poverty level of the noncustodial parent, custodial family, and married parent family. The charts in the left column show poverty level in the absence of child support—the right column shows the poverty level after child support. The three rows show poverty level before federal income and payroll taxes and the EITC, after taxes but before the EITC, and after taxes and the EITC. As in table 1, child support is subtracted from the income of the noncustodial parent and added to the income of the custodial family when calculating the poverty level.

The top left, middle right, and bottom right charts in table 2 illustrate the figures shown in table 1. The remaining charts show poverty levels based on other combinations of child support and pre and post-tax income. These charts make possible some additional observations. For example, moving from left to right in the top row, we see that child support lowers the noncustodial parent's pre-tax poverty level from 147 percent to 122 percent, while raising the custodial family's from 114 percent to 133 percent. Moving from the top to middle chart in the left column, we see that taxes prior to EITC lower the poverty level of both the noncustodial parent and the custodial family. Although taxes have the biggest effect on the poverty level of the noncustodial parent, the noncustodial parent's poverty level (126 percent) remains substantially higher than that of the custodial family (104 percent). From this point, adding child support raises the custodial family's poverty level to 122 percent (middle right chart). Adding the EITC, but not child support, raises the custodial family's poverty level to 123 percent (bottom left chart). Adding both the EITC and child support raises the custodial family's poverty level to 142 percent (bottom right chart)—while paying child support and taxes lowers the noncustodial parent to 101 percent of poverty.

Tables 1 and 2 illustrated the case of low income parents with one child.Table 3 shows what happens to poverty levels when a second child is added. Adding a second child has several effects (not shown): the custodial parent gets an additional dependent exemption which eliminates her remaining $150 in federal income taxes; the custodial parent becomes eligible for a higher EITC benefit ($3,482 instead of $2,096); the noncustodial parent's child support obligation increases from $2,001 to $3,006; and the custodial family's poverty threshold increases from $10,562 to $12,517.

By comparing tables 2 and 3 we can see the effect of these changes. First, the addition of the second child substantially lowers the standard of living of the married parent family. As a result, divorce has less of an effect on the noncustodial parent's poverty level than when there was only one child. In each case, the noncustodial parent's poverty level (after taxes and EITC) is 126 percent. However, this is relative to 167 percent of poverty for a married couple family with one child, but only 138 percent of poverty for a married couple family with two children. Second, in the absence of EITC and child support, the additional child drives the custodial family into poverty. Again, child support and EITC have approximately equal effects on the custodial family's poverty level—in this case raising the family to 113 percent and 116 percent of poverty respectively. Interestingly, the combined effect of child support and EITC makes the custodial family with two children almost as well off as if there was only one child (141 percent and 142 percent of poverty respectively)—and raises the custodial family above the poverty level of the married couple family with two children (138 percent). Paying child support for the second child impoverishes the custodial parent—driving him from 101 to 88 percent of poverty after taxes and child support.

Tables 4 (for one child) and 5 (for two children) extend this analysis to different combinations of noncustodial parent and custodial parent earnings. These tables show poverty levels for the noncustodial parent, custodial family, and married parent family after child support, federal income and payroll taxes, and the EITC. The rows and columns show poverty levels at earnings of $9,000, $12,000, $15,000, and $18,000 for the custodial and noncustodial parent respectively. Earnings below $9,000 are omitted because states tend to award only token amounts of support to noncustodial parents below poverty. Also, custodial families with earnings below $9,000 are more likely to rely on welfare, making an analysis that focuses only on earnings unrealistic. At the opposite extreme, charts in the lower right corner depict families that might be described as middle income. Indeed, married parent families with one child easily exceed 200 percent of poverty. We include these charts for completeness.

Looking first at table 4, we see that at almost all of the combinations of noncustodial and custodial parent earnings shown, the custodial family fares better than the noncustodial parent after federal income and payroll taxes, the EITC, and child support. In these income ranges, the noncustodial parent must earn roughly 50 percent more than the custodial family to have the same standard of living (as in the case of a noncustodial parent with $15,000 and a custodial parent with $9,000). Table 5 shows that even an additional 50 percent is not enough when there are two children—the noncustodial parent must earn twice as much ($18,000) as the custodial parent ($9,000) to have an equal standard of living. In addition, table 5 shows that low income custodial families with two children (after taxes, EITC, and child support) are only slightly worse off, and in many cases better off, than had the two parents been married.

The examples in tables 1 through 5 oversimplify in a number of ways. They ignore the effect on standard of living of means tested government assistance, such as welfare and food stamps, which can be expected to disproportionately assist the custodial family. They also ignore the costs of child care and health insurance.7 Finally, the examples do not take into account the non-reimbursed labor of rearing children, nor the psychological benefits of having children present. These simplifications become less relevant, however, when considering this as a model of a world in which custodial parents work rather than rely on welfare, and in which child support guidelines distribute the burden of health insurance and child care costs between the two parents. In this world, low income, child support paying, noncustodial parents fare worse than custodial families, unless the noncustodial parent earns approximately 50 to 100 percent more than the custodial parent.

Our findings are sensitive to the fact that we chose to measure relative well-being using the official poverty thresholds rather than some other scale. While the official poverty thresholds have the advantage of being familiar and widely used, they are not without problems. As Patricia Ruggles (1990) discusses, the methods used to adjust the poverty threshold for family size were developed in an ad hoc manner and are extremely outdated. Adjustments for family size comprise an "equivalence scale" that indicates how much more money a family of each size must have, relative to a single individual, to have the same standard of living as the single individual. Under the official poverty thresholds, a family of two requires only 29 percent more than a single individual in order to have the same standard of living as the single individual. The two member family requires less than twice as much income as the single individual because of "economies of scale"—most notably, shared housing expenses, transportation, and utilities. As Ruggles discusses, a number of other estimates of equivalence scales have been made over the years, with the official poverty thresholds implying greater economies of scale (and therefore smaller additional needs per family member) than most other estimates.

In 1995, the National Academy of Sciences (NAS) issued an exhaustive review of poverty measurement (Citro and Michael 1995). One of the many recommendations of this report was that the equivalence scale be set to treat children as consuming 70 percent as much as adults on average, and that economies of scale should be computed by raising the number of adult equivalents in a family to a power of between .65 and .75. To see the effect that such an equivalence scale would have on our results, we created hypothetical poverty thresholds by applying the NAS equivalence scale to the official 1996 poverty threshold for a single adult. Because we use the official poverty threshold for single adults, our estimates of the poverty status of the noncustodial parent do not change. However, custodial and married couple families appear worse off under the hypothetical thresholds than under the official poverty thresholds. For example, if each parent earned $12,000 and child support was paid on behalf of one child, the custodial family would have been between 124 percent and 130 percent of poverty under the hypothetical thresholds, as compared to 142 percent of poverty under the official poverty thresholds. We would still conclude that the noncustodial parent fares worse than the custodial parent (who would remain at 101 percent of poverty) but the difference would be less extreme, and less would be required (20 to 50 percent additional earnings for the noncustodial parent) to raise the noncustodial parent to the same standard of living as the custodial family.

Obviously, the choice of equivalence scale is critical to an analysis such as this. Since the official poverty thresholds imply greater economies of scale than most other equivalence scales, our results are best viewed as an upper bound estimate of the discrepancy in the standards of living of low income noncustodial and custodial parents.

Who Should Bear the Cost of Marital Dissolution and Non-Marital Birth?

The examples above show that the combined effect of federal income taxes, the EITC, and typical child support awards is a lower standard of living for low-income noncustodial parents than for custodial families with similar earnings. Only when the earnings of the noncustodial parent exceed those of the custodial family by as much as 50 percent (for one child) or 100 percent (for two children) does the low-income noncustodial parent fare as well as the custodial family.

Given a society which favors limits on government assistance to the poor while seeking to protect the well-being of children, it is not inconsistent that low-income noncustodial parents be asked to bear the brunt of the cost of marital dissolution and non-marital birth. While it would be ideal for custodial and noncustodial parents to bear the cost equally, it is not possible to lower the standard of living of the custodial parent without also lowering the standard of living of the children.

On the other hand, current guidelines, coupled with federal income tax law, leave low-income noncustodial fathers with little after paying taxes and child support. These parents are helping to support children, but under current tax law receive little of the tax relief available to families with children. Low income noncustodial parents could be made better off through two strategies: changing child support guidelines to take greater account of the effect of taxes on the well-being of noncustodial and custodial families, or changing the tax law to extend some of the tax relief available to families with children to noncustodial parents who pay child support. In this paper, we explore the latter strategy. Making relief contingent on payment of child support provides an additional incentive to pay child support. Our chosen method for providing relief, an expansion to the earned income tax credit, has the additional advantage that it helps to encourage work. Furthermore, it is targeted toward low-income noncustodial parents, whose standard of living is most at risk under current guidelines and tax law.


Section II: The EITC

In this section we provide further detail about the EITC. We begin with a brief legislative history of the EITC. We then discuss 1996 eligibility criteria and benefit levels and the work incentives and disincentives generated by the EITC. Using data from the 1990 Survey of Income and Program Participation (SIPP) we examine the characteristics of persons who would have been eligible for the EITC had the 1996 eligibility rules and benefit levels been in effect in 1990. Finally, we narrow our focus to the characteristics of custodial mothers and noncustodial fathers eligible for the EITC.

Legislative History of the EITC

Until 1975, the federal government directed cash assistance for the poor to certain categories of people whose poverty largely resulted from the absence of income—the aged, the disabled, and children in families with an absent parent. Although this approach to helping the poor directed dollars toward those considered most deserving, it created disincentives to work. Furthermore, it did not provide relief to those who worked and remained poor. Starting in the early 1970s, Congress began to consider ways to help the working poor.

In the midst of a major recession in 1975, Congress passed the Tax Reduction Act of 1975, which included a small one-year refundable credit to low-income working families. The rationale given for this provision was to provide relief to low-income families and help offset the impact of Social Security taxes on low-income workers. It was also seen as an alternative to welfare and an incentive to work. The EITC was extended, through annual legislative tax provisions, from 1975 through 1977, and made permanent (and more generous) in 1978. It was increased in 1984 and then again in 1986, when it was also indexed to keep pace with inflation.

In 1990 the EITC underwent a major expansion. Under the Omnibus Budget Reconciliation Act (OBRA) of 1990, Congress not only increased the maximum credit and level of income that eligibles could earn, it also created a larger EITC benefit for families with two or more children. Prior to 1990, families with two or more children received the same EITC benefit as families with one child. In addition, Congress enacted two changes that benefited recipients of means-tested programs. First, it instructed other federally funded means-tested programs not to count EITC benefits as income in determining eligibility or benefit amounts. Prior to this law, most means-tested programs were supposed to treat EITC benefits as income. Second, Congress changed the eligibility criteria for EITC, making it simpler for welfare recipients to qualify for the EITC. Prior to OBRA 1990, a person was eligible for the EITC if he/she lived with a child more than half the year and provided over half the child's total support. OBRA 1990 eliminated this support test as a criterion for eligibility.

In 1993, Congress enacted another major expansion of the EITC. It increased the credit for one- and two-child families and extended the credit for the first time to low-income working adults with no qualifying children. The main rationale for this latter credit, called the childless credit, was to partly offset the effect of increased consumption and payroll taxes on low-income workers and to provide an incentive to work.

Eligibility Criteria, Benefit Levels, and Work Incentives and Disincentives

The EITC is available to families who meet certain demographic requirements. Eligible claimants fall into three groups: (1) childless adults between the ages of 25 and 64; (2) individuals with one qualifying child; and (3) individuals with two or more qualifying children.

To claim a qualifying child, three tests must be satisfied: (1) a relationship test—-the child must be the son, daughter, step-child, grandchild, foster child, or adopted child of the claimant; (2) an age test—the child must be under 19, or under 24 and a full-time student for at least 5 months, or permanently and totally disabled; and (3) a residency test—the child must live with the claimant for more than half the year.

Tax filers eligible for the EITC must also meet certain income requirements. First, they must have earned income. Second, their earned income and Adjusted Gross Income (AGI) must be within a certain range (see Table 1). Third, their investment income must be no greater than $2,200. Investment income includes interest income, dividends, net capital gains, rent and royalties (that are from sources other than the filer's ordinary business activity), and net passive income. Paying or receiving child support has no effect on the amount of the EITC credit. This treatment of child support is consistent with the treatment of child support within the income tax system—child support payments are taxed as income for the noncustodial parent but not for the custodial parent.

Unlike traditional welfare programs, the EITC increases as earnings increase, as long as earned income falls within a phase in range. Once the maximum credit is reached, there is a plateau range, within which the credit does not change as earnings rise. At a certain income level, the credit begins to decline as income rises. This is the phase outrange. Eventually the credit phases out altogether.

Table 6 describes the EITC parameters in 1996. The table shows that a family with two or more children receives EITC equal to 40 percent of earnings of $8,910 or less, for a maximum annual credit of $3,556. Families with one child receive EITC equal to 34 percent of earnings of $6,330 or less, for a maximum credit of $2,152. The credit begins to phase out at $11,610 of earnings (or AGI if higher) for both family types. For families with two or more children, the credit is phased out at a rate of 21.06 cents for each dollar of earnings (or AGI if higher) in excess of $11,610, causing the credit to be completely phased out at $28,495. The phase out rate for families with one child is 15.98 cents per additional dollar, causing the credit to be completely phased out at $25,078.

The EITC for childless workers is much smaller—only 7.65 of earnings up to $4,220. This exactly offsets the worker share of Social Security and Medicare payroll taxes on earnings below this amount. The credit is phased out beginning at earnings (or AGI) of $5,280 and is completely phased out at $9,500.

One of the primary goals of the EITC is to provide an incentive to work. Because the EITC increases income available through work, it encourages work as an alternative to welfare. Also, the phase in of the credit increases the marginal wage of recipients in the phase in range. However, for those who would work in the absence of the EITC, the EITC provides incentives to reduce work effort. The additional income from the EITC makes it possible for families to have the same level of income as they would have in the absence of the EITC while working fewer hours. Furthermore, the phase out of the EITC increases the marginal tax rate of those in the phase out range. Efforts to improve financial work incentives for those currently outside the workforce through a program such as the EITC necessarily imply increasing the marginal tax on work as the credit is phased out at higher income ranges. Despite incentives to reduce hours of work, most analysts believe that the overall effect of the EITC is to increase total hours of work in the low income population.

Characteristics of EITC Eligibles

Using the 1990 Survey of Income and Program Participation (SIPP), we estimate that 16.3 million tax filers would have been eligible for the EITC if the 1996 EITC rules had been in effect at that time. (See the Appendix for a discussion about the difference between our estimated number of eligibles and the number of EITC recipients.) Nearly 5 million adults would have been eligible for the EITC childless credit and 11.6 million families with children would have been eligible for the EITC qualifying child credit (see Table 7). The average benefit for those receiving the childless credit would have been $167 (in 1996 dollars) and the total cost would have been $0.8 billion. Those receiving the qualifying child credit would have received an average benefit of $1,596 for a total cost $18.0 billion.

Table 7 shows that the 1996 EITC rules are not particularly well targeted toward the poor. Only 33 percent of the 1996 EITC benefits would have gone to the poor in 1990 had those rules been in effect at that time. Half of the benefits, however, would have gone to families with incomes between 100 and 200 percent of the poverty threshold for their family size. The final 17 percent would have gone to families with incomes above 200 percent of the poverty threshold. More of the benefits under the childless credit would have gone to the poor than under the qualifying child credit, but nonetheless the majority of these EITC dollars would have also gone to individuals with incomes above the poverty threshold.

There are two primary reasons why such a low proportion of EITC benefits go to the poor. First, for families with children, the EITC is not completely phased out until income levels well in excess of 200 percent of poverty. As a result, only 26 percent of EITC eligibles with a qualifying child have incomes below poverty. Of those eligible for the childless credit, 37 percent are poor. The fact that so few of those eligible for the childless credit are poor may seem surprising at first, since the point at which the credit is phased out ($9,500) is only sixteen percent higher than the poverty threshold for one person ($8,163). However, many of these eligibles live with relatives, whose additional income raises them above poverty.

Characteristics of Eligible Custodial Mothers and Noncustodial Fathers

EITC recipients include a diverse group of people—married couples with or without children, single and married custodial parents with children, noncustodial parents without qualifying children, noncustodial parents who also have qualifying children, and single individuals without children. In Table 8, we restrict our focus to custodial mothers with children and noncustodial fathers—including those who also have one or more qualifying children. Although there are a growing number of cases in which it is the father who is the custodial parent, our data restrict us to analyzing the more typical case of the custodial mother and noncustodial father.

Table 8 shows that 4.7 million custodial mothers would have been eligible for an average of $1,644 in EITC benefits in 1990, had the 1996 EITC rules been in effect. Poor custodial mothers would have received 29 percent of total EITC benefits; half would have gone to mothers with incomes between 100 and 200 percent of the poverty threshold; and 22 percent would have gone to families above 200 percent of the poverty threshold. Nearly half of custodial mothers eligible for the EITC received child support. Of those receiving child support, an average of $3,153 was received.

As for noncustodial fathers, 1.3 million would have been eligible for the qualifying child credit and 0.8 million for the childless credit. Noncustodial fathers with a qualifying child would have been eligible for an average of $1,602, compared to only $185 for those without a qualifying child. A somewhat higher proportion of EITC expenditures for noncustodial fathers would have gone to the poor than is the case for custodial mothers, but nonetheless most would have been paid to the nonpoor. Few noncustodial fathers eligible for the EITC pay child support—only 27 percent of those with a qualifying child and 28 percent of those eligible for the childless credit. Of those paying child support, the average amount paid by those eligible for the childless credit ($1,625) equals 23 percent of their average income, compared to only 14 percent ($2,284) for those with a qualifying child.


Section III: Policy Options and Financing Methods

In section I, we described the effect of typical child support guidelines and current tax law on low income noncustodial parents and custodial families. We found that low income noncustodial parents require earnings 50 to 100 percent higher than the custodial family in order to pay child support and enjoy the same standard of living as the custodial family. In section II, we showed that while the EITC provides an important benefit for custodial mothers, fewer than half as many noncustodial fathers benefit. While the average benefit for noncustodial fathers with a qualifying child is about the same as for custodial mothers, the average benefit for the childless credit is much smaller—only $185.

In this section, we explore three options for expanding the EITC to greater assist noncustodial parents who pay child support and discuss two options for partially or fully financing these options. Our three policy options involve waving the residency test (i.e., that the "qualifying" child must live with the parent more than half of the year) for noncustodial parents who meet their full child support obligation. The three options could be partially or fully financed by two different methods that reduce EITC benefits for custodial families receiving child support. After discussing the policy options and financing methods, we consider the effect of two different policy/financing combinations on the relative well being of noncustodial parents and custodial families.

Three Options for Expanding the EITC

The first policy option would extend the EITC/qualifying child credit to noncustodial parents who meet their full child support obligation, so long as they also meet the relevant income criteria. Thus, noncustodial parents who pay child support for one child would be eligible for the EITC as if they had one qualifying child; noncustodial parents who pay child support for two or more children would be eligible for the EITC as if they had two or more qualifying children. The credit would be capped, however, at the amount of child support paid. For example, the EITC credit for a family with $12,000 and one qualifying child is $2,096. In the example shown in table 1, a noncustodial parent with one child pays $2,001 in child support. Under this policy option, the parent would be capped at $2,001 in EITC benefits, and so would not receive the full $2,096 available to a parent with a qualifying child.

The second policy option is similar to the first, but instead of allowing both parents to be eligible for the EITC if they meet the income criteria, only one parent would be eligible. Under current law, only the custodial parent can be eligible for the EITC. Under this policy option, a noncustodial parent would be eligible for the qualifying child credit if he or she meets the relevant income criteria, pays the full amount of child support due, and if the custodial parent is not eligible for the EITC. In other words, the custodial parent would be given priority for the EITC, but if she or he is ineligible, the qualifying child credit would be extended to the noncustodial parent.

The third policy option gives noncustodial parents who pay child support half of the credit amount for which they would be eligible if their children lived with them, or the amount of child support paid, whichever is lower. For example, a noncustodial parent earning $12,000 and paying $2,001 for one child would receive an EITC benefit of $1,048 (one half of the EITC amount for one child). This option is based on the premise that noncustodial parents who pay child support in full are paying approximately half of the child's support and thus should be entitled to half of the earned income credit.

Costs and Benefits Under Each Policy Option

For each policy option, table 9 reports the number of noncustodial fathers who would benefit, the average increase in their EITC, and the amount by which aggregate EITC costs would increase. The first policy option would increase EITC benefits by an average of $1,286 for approximately 747 thousand noncustodial fathers. Aggregate EITC costs would increase by 5.1 percent. Under the third policy option, about the same number of noncustodial fathers would benefit, but the average increase would be $706. Aggregate EITC costs would increase by 2.8 percent.

The second policy option is more difficult to estimate because noncustodial fathers would only be eligible in cases in which the custodial mother does not claim the EITC. Since the SIPP does not link noncustodial fathers with corresponding custodial mothers, we must estimate the number of noncustodial fathers who would be eligible for the EITC under this option. To do this, we count the number of custodial mothers who receive child support in full and have AGI within the range covered by the EITC, but are ineligible for the EITC due to lack of earnings. If we assume a one-to-one correspondence between these custodial mothers and the number of noncustodial fathers eligible under this policy option, then about 349 thousand noncustodial fathers would be eligible.8 Assuming these fathers receive an additional average $1,286 in EITC benefits (as under policy option 1), aggregate EITC costs would increase by 2.4 percent.

Table 9 shows that none of these policy options is particularly well targeted toward the poor. Only 2 to 3 percent of the additional EITC dollars generated from these expansions would go to the poor. About half of the benefits would go to families with incomes above 200 percent of the poverty threshold. Nonetheless, the expansion would help fathers who currently spend an average of $3,450—or 20 percent of their income—on child support.

The costs of these policy options may be larger than estimated if low income noncustodial fathers respond by paying more child support. On the other hand, additional child support generated by these policy options would undoubtedly benefit children and help reduce welfare costs.

Financing the Expansion

We examine two possible methods for partially or fully financing these policy options. The first method would be to count child support exceeding a certain dollar amount when calculating the custodial parent's EITC. For example, if a custodial parent received more than $2,400 in child support for one child or $3,600 for two or more children, the additional amount would be added to her AGI (or earnings, if higher) in phasing out the EITC. Currently, a custodial parent could receive any amount of child support and still be eligible for the EITC, since child support income is not considered in calculating the EITC benefit.9 Since custodial parents with high amounts of child support tend to be better off, virtually all of the savings from this financing method would be made at the expense of non-poor custodial families. Only a fraction of the savings (0.2 percent) would be made at the expense of the poor, whereas 2 to 3 percent of the benefits generated by the aforementioned expansions would go to the poor. Overall, this financing method would reduce EITC costs by 1.2 percent—about half of that necessary to finance the second policy option and a quarter of that necessary to finance the first policy option.

The second financing option would be to reduce EITC benefits by half for custodial parents who receive their full child support obligation. Aggregate EITC costs would be reduced by 4.7 percent, creating savings that exceed the cost of extending one half of the qualifying child EITC to noncustodial parents (i.e., policy option 3 above). Savings from this financing method exceed the costs of policy option 3 because, within the income range covered by the EITC, there are more custodial mothers who receive, than noncustodial fathers who pay, the full amount of child support due. However, this financing option and the third policy option are not well targeted: 18 percent of the cost savings would come at the expense of poor custodial mothers, but only three percent of the benefits would go to poor noncustodial fathers.

Combined Effects of Policy and Financing Options on Relative Well-Being

In Table 10 we examine the effect of two policy and financing combinations on the relative standards of living of a custodial family and noncustodial parent, assuming each has $12,000 in earnings and the noncustodial parent pays child support according to typical income shares guidelines. The left column shows the results for one child and the right column shows the results for two. The poverty level (under current law) of the noncustodial parent and custodial family after taxes and child support is shown in the top row. As before, we show the after tax poverty level of a married parent family for reference.

Under the first policy option, noncustodial parents who pay child support for one child would be eligible for the EITC as if they had one qualifying child; and noncustodial parents who pay child support for two or more children would be eligible for the EITC as if they had two or more qualifying children. The credit would be capped, however, at the amount of child support paid. Under the first financing method, child support in excess of $2,400 (for one child) or $3,600 (for two or more children) would be counted in phasing out the custodial parent's EITC. The combined effect of this policy option and financing method is shown in the second row of table 10. In this case, the custodial family is unaffected, because the amount of child support received ($2,001 for one child and $3,066 for two children) is less than the amount at which child support begins to be counted in calculating the EITC phase out ($2,400 for one child and $3,600 for two children). However, this policy option has a major impact on the noncustodial parent—lifting him from 101 percent of poverty (88 percent of poverty for two children) to 126 percent of poverty. Although benefiting from this policy option, the standard of living of the noncustodial parent remains somewhat lower than that of the custodial family.

Expanding this analysis to the full range of noncustodial and custodial parent incomes between $9,000 and $18,000 (not shown) we find that the effect of this policy/financing combination on custodial families is virtually non-existent, while noncustodial parents experience a substantial increase in standard of living, especially at the lower income levels. Although noncustodial parents remain worse off than custodial families with equal earnings, the extent of discrepancy is reduced. As a result, noncustodial parents require only about 20 to 30 percent more earnings than the custodial family to be equally well off, compared to 50 to 100 percent under current law. As would be expected, this policy option has a bigger effect on the poverty level of noncustodial parents with two children than of noncustodial parents with one child, since the additional child implies a higher child support amount and EITC credit.

Under the third policy option, noncustodial parents paying the full child support amount would be eligible for half of the qualifying child EITC, capped at the amount of child support paid. The second financing method allows custodial parents receiving the full amount of child support only half of the qualifying child EITC for which they would otherwise be eligible. The combined effect of this policy option and financing method (as shown in the third row of table 10) is to raise the standard of living of the noncustodial parent (although by not as much as in the first policy option) and to lower the standard of living of the custodial family. Although the noncustodial parent in this example would remain worse off than the custodial family, the difference in the two standards of living would be reduced. The poverty level of the noncustodial parent would rise from 101 to 114 percent (or from 88 to 110 percent, for two children), relative to current law, and the poverty level of the custodial family would fall from 142 to 132 percent (or from 141 to 127 percent, for two children).

Expanding this analysis to the full range of noncustodial and custodial parent incomes between $9,000 and $18,000 (not shown), we find that the effect of the third policy option and second financing method is a higher standard of living for the noncustodial parent and a lower standard of living for the custodial family, especially at the lowest income levels. Like the first policy option and financing method the discrepancy in the standard of living of the noncustodial parent and custodial family is reduced. However, unlike the first policy option and financing method, this reduction involves lowering the standard of living of custodial families, especially at the lowest income levels.


Section IV: Behavioral Effects

The policy options and financing methods discussed above affect incentives involving the payment and receipt of child support, the decision of whether and how much to work, and the decision of whether or not to marry. If enacted, they would also affect the extent of noncompliance with EITC rules. In this section, we consider each of these issues.

Incentives for Payment of Child Support

Because eligibility would be contingent upon full payment of the child support award, the EITC expansion would provide an incentive to pay the full amount due. Depending on the policy option, child support award level, and noncustodial parent's income, some or all of the child support paid would be reimbursed to the noncustodial parent through the EITC. Because the amount of the EITC is capped at the amount of child support paid, it enables noncustodial parents to at best break even—they are not made better off than they would have been if they had not paid child support. Thus, we expect these provisions to provide the strongest incentives to noncustodial parents who currently pay only part of what is due, or who pay child support informally. Those who currently make partial payments might find themselves capable of paying the full amount—if they were assured of partial or full reimbursement through the EITC. Those who assist their children informally would have an incentive to enter into a formal award or agreement in order to become eligible. Noncustodial parents who provide no financial assistance to their children might or might not respond to incentives generated by the policy options. If the noncustodial parent has no desire (whether legally or morally motivated) to provide for his children, the policy options are unlikely to have any effect. On the other hand, if the noncustodial parent feels some motivation to provide for his children, these policy options might be enough to tip the balance in favor of paying child support.

Incentives for Receipt of Child Support

Under the second financing option, custodial families receiving the full amount of child support would have EITC reduced by half. Under this scenario, if child support is less than half of the EITC, a custodial family would be better off without the child support. In most cases, this will not be a problem—child support awards typically exceed one half of the maximum earned income credit. However, in many states, noncustodial parents below the poverty threshold are charged a token amount of support, for example $25 or $50 per month. The resulting $300 or $600 in annual child support is easily less than half of the EITC for many recipients. These custodial families would have a clear incentive not to receive child support, running directly counter to efforts to encourage low income custodial families to seek child support. This problem could be avoided by modifying the second financing option to reduce EITC by no more than the amount of child support received.

With the exception of the case described above, most custodial mothers would continue to be better off receiving the full amount of child support due than receiving no child support. However, since EITC would be reduced by half only if the custodial parent received the full amount of child support, there would be a strong incentive for custodial parents to encourage noncustodial parents to pay less than the full amount. Of course, by not paying the full amount, noncustodial parents would forfeit their own right to the qualifying child EITC. However, this would not affect noncustodial parents whose incomes make them ineligible for the EITC. In these cases, both the (ineligible) noncustodial parent and the custodial family would be better off if less than the full amount of child support was paid, so long as the amount lost did not exceed one half of the custodial family's EITC.

Incentives Affecting Work: Noncustodial Parents

By using the EITC to encourage payment of child support, our policy options would expose many more noncustodial parents to EITC work incentives. For nonworkers for whom working would mean paying child support (as in the case of automatic wage withholding) the promise of additional income from the EITC qualifying child credit might make work seem worthwhile. Noncustodial parents who already work and pay child support might reduce hours of work—since the additional income from the EITC would allow them to have the same standard of living while working fewer hours. Noncustodial parents in the phase in range would enjoy a higher marginal wage rate, while those in the phase out range would experience a higher marginal tax rate. However, for those whose child support is lower than the maximum credit, the phase in and phase out ranges would be reduced—phase in would occur only up to the point at which the EITC is equal to the award, and phase out would occur only after the EITC credit is reduced to an amount equal to the award.

Incentives Affecting Work: Custodial Parents

The policy options might also affect work incentives for custodial parents. Additional child support income generated in response to the policy options might be enough to prompt nonworkers to choose child support and work over reliance on welfare, and might enable working custodial parents to reduce hours of work. On the other hand, the second policy option might encourage noncustodial parents to pressure custodial parents not to work so that the noncustodial parent would be eligible for the EITC.

By reducing the amount of EITC available at each income level, the second financing method would reduce existing incentives to work fewer hours. The first financing method would make some recipients ineligible for the EITC and would make others eligible for a smaller amount, also reducing incentives to work fewer hours. On the other hand, this method might move some custodial parents into the phase out range, increasing their marginal tax rate and thus increasing incentives to work fewer hours.

Incentives Affecting Marriage

Under current law, the EITC exacerbates the "marriage penalty" that exists for two-earner families. For example, if each parent earns $15,000 per year, their combined incomes make them ineligible for the EITC. However, if they divorce, the parent with custody of the children becomes eligible for an EITC benefit of $3,000, while the noncustodial parent receives no EITC benefits.

In contrast, the EITC gives one-earner couples an incentive to marry. For example, if a mother stays home while the father earns $15,000, this family (with two children) would be eligible for a $3,000 credit, but if they separate and the mother has custody of the children while the father pays $4,500 in child support, neither parent would be eligible for the EITC.

All three policy options to extend the EITC would exacerbate the EITC's "marriage penalty" because they would allow both parents in a divorced working couple to receive a tax credit that neither would have been eligible for if married. However, financing options that reduce custodial families' EITC would mitigate this effect somewhat. In addition to exacerbating the EITC's marriage penalty, the three policy options reduce the EITC's current marriage incentive, because noncustodial fathers who pay child support would be eligible for a tax credit for which they are ineligible under current law.

Compliance with EITC Rules

The EITC expansion might actually increase compliance with EITC rules by legitimizing an area of prior noncompliance (noncustodial parents claiming the EITC).10 Of course, new complexities would be introduced—in particular, validating the amount of the child support award and whether the full amount was paid. Under the second policy option, there would be the additional complexity of ensuring that only one parent claimed the qualifying child EITC. However, the creation of a national registry of child support orders (mandated under the Personal Responsibility and Work Opportunity Reconciliation Act of 1996) provides the possibility for greater integration of the tax and child support systems, and a potential mechanism for verifying child support related EITC claims.


Summary and Conclusions

In this paper we show that the combined effect of federal income tax law and child support paid under typical state guidelines is a lower standard of living for the noncustodial parent relative to a custodial family with similar earnings. Only when the earnings of the noncustodial parent exceed those of the custodial parent by 50 to 100 percent does the noncustodial parent enjoy the same standard of living as the custodial family. Despite helping to support their children, noncustodial parents receive little of the tax relief available to families with children.

We have explored three options for extending the qualifying child EITC to noncustodial parents who pay child support: 1) allowing eligibility for the full qualifying child credit; 2) allowing eligibility for the credit if the custodial parent is ineligible; and 3) allowing eligibility for half of the credit. In each of these cases the amount of EITC received would be capped at the amount of the child support award. We estimate that these policy options would increase EITC expenditures from 2.4 to 5.1 percent. These estimates do not reflect the additional cost generated by noncustodial parents who pay child support as a result of the policy change, but neither do they reflect the impact of the additional child support on the poverty and welfare status of poor custodial families.

Our first financing option, counting child support in excess of $2,400 ($3,600 for two or more children) in phasing out the EITC, would save enough money to fund one quarter of the first policy option or one half of the second policy option. Furthermore, only a fraction of the savings (0.2 percent) would be made at the expense of the poor, whereas 2 percent of the benefits would go to the poor.

The second financing option (reducing the EITC by half for custodial parents receiving child support in full) would produce more than enough savings necessary to finance the third policy option (extending half of the qualifying child EITC to noncustodial parents who pay child support in full.) However, this policy and financing option is poorly targeted: 18 percent of the cost savings would come at the expense of poor custodial mothers, but only three percent of the benefits would go to poor noncustodial fathers. Furthermore, this financing option creates incentives for custodial parents not to receive the full amount of child support due. Therefore, we recommend against this policy option.

Unfortunately, rejecting the second financing option means finding other sources of funding for the expansions proposed here. Savings from the first financing method could fund from one quarter to half the cost, depending on the policy option. Additional savings would be available if better integration of tax and child support systems improves tax compliance (relating to the dependent exemption and EITC) among noncustodial parents. Essentially, this would transfer money from ineligible noncustodial parents who do not pay child support to eligible custodial parents who do. Finally, to the extent that noncustodial parents who pay child support are currently (inaccurately) claiming the qualifying child EITC, we overestimate the cost of expansion.

Even if funding must be sought elsewhere, there are legitimate reasons to seek assistance for low income noncustodial fathers through the EITC. Typical state child support guidelines, coupled with federal income tax law, leave low-income noncustodial fathers with little after paying taxes and child support. These parents are helping to support children, but under current tax law receive little of the tax relief available to families with children. As a result, a low income noncustodial parent who pays child support must earn 50 to 100 percent more than the custodial parent in order to enjoy the same standard of living. Extending partial or full eligibility for the qualifying child EITC to these parents would help to reduce this inequity, and would provide an incentive for other noncustodial parents to work and pay child support.


References

Bassi, Laurie J., Laudan Aron, Burt S. Barnow, and Abhay Pande. 1990. "Estimates of Expenditures on Children and Child Support Guidelines." Report submitted to Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. Washington D.C.: Lewin/ICF.

Bassi, Laurie J. and Burt S. Barnow. 1993. "Expenditures on Children and Child Support Guidelines." Journal of Policy Analysis and Management (Volume 12, Number 3) pp. 478-497.

Citro, Constance F. and Robert T. Michael (eds). 1995. Measuring Poverty: A New Approach Washington, D.C.: National Academy Press.

Duncan, Greg J. And Saul D. Hoffman. 1985. "A Reconsideration of the Economic Consequences of Marital Dissolution." Demography (Volume 22, Number 4) pp. 485-497.

Gabe, Thomas. 1993. "The Earned Income Tax Credit Current Law and the Clinton Proposal: Characteristics of Eligible Families." Congressional Research Service.

Holden, Karen C. And Pamela J. Smock. 1991. "The Economic Costs of Marital Dissolution: Why do Women Bear a Disproportionate Cost?." Annual Review of Sociology (Volume 17) pp. 51-78.

Internal Revenue Service. 1997. "Study of EITC Filers for Tax Year 1994." (April).

Ruggles, Patricia. 1990.Drawing the Line: Alternative Poverty Measures and Their Implications for Public Policy Washington, D.C.: The Urban Institute Press.

Scholz, John Karl. 1994. "The Earned Income Tax Credit: Participation, Compliance, and Antipoverty Effectiveness."National Tax Journal (March) pp. 63-87.

Smock, Pamela J. 1994. "Gender and the Short-Run Economic Consequences of Marital Disruption." Social Forces (September) pp. 243-262.

Sorensen, Elaine. 1997. "A National Profile of Nonresident Fathers and their Ability to Pay Child Support."Journal of Marriage and the Family (November): forthcoming.

Sorenson, Annemette. 1992. "Estimating the Economic Consequences of Separation and Divorce: A Cautionary Tale from the United States." InEconomic Consequences of Divorce, ed. Lenore J. Wietzman and Mavis Maclean, pp. 263-282. New York: Oxford University Press.

Stirling, Kate J. 1989. "Women Who Remain Divorced: The Long-Term Economic Consequences." Social Science Quarterly (Volume 70, Number 3) pp. 549-561.

Storey, James R. 1996. "The Earned Income Tax Credit: A Growing Form of Aid to Low-Income Workers." Congressional Research Service.

U.S. Bureau of the Census. 1991.Child Support and Alimony: 1989.Current Population Reports, Consumer Income, Series P-60, No. 173. Washington, D.C.

U.S. Bureau of the Census. 1993.Statistical Abstract of the United States: 1993 (113th edition) Washington, D.C.

U.S. Department of Health and Human Services. Office of Inspector General. 1996. "Unpaid Child Support and Income Tax Deductions."

U.S. General Accounting Office. 1997. "Earned Income Credit Noncompliance." GAO/T-GGD-97-105 (May).

U.S. General Accounting Office. 1996. "Earned Income Credit: Profile of Tax Year 1994 Credit Recipients." GAO/GGD-96-122BR (June).

Weiss, Robert S. 1984. "The Impact of Marital Dissolution on Income and Consumption in Single-parent Households."Journal of Marriage and the Family(February) pp. 115-127.

Weitzman, Lenore J. 1985.The Divorce Revolution: The Unexpected Social and Economic Consequences for Women and Children in AmericaNew York: The Free Press.


Appendix

This analysis uses the 1990 Survey of Income and Program Participation (SIPP), a large nationally representative survey conducted by the U.S. Census Bureau. Approximately 23,000 households were interviewed in 1990 and were asked detailed questions about their income, earnings, and family situations. Although the SIPP is an excellent data source for determining EITC eligibility, there are four important caveats to remember for this analysis. First, the data have been reweighted to reflect the underrepresentation of noncustodial fathers in the SIPP (see Sorensen (1997) for a detailed explanation). Second, although we use 1996 EITC rules (and 1994 rules in the validation exercise below) the data are from 1990. Thus, the estimates reflect population size and characteristics in 1990. Third, although the 1990 SIPP has most of the information needed to calculate EITC eligibility, it does not include all information needed. Thus, we adopt simplifying assumptions which are described in greater detail below.

The fourth important caveat is that we describe the eligible EITC population, not the actual recipient population. Recipients and eligibles are expected to differ in important ways. First, recipients consist of two groups—those who are eligible and those who are ineligible. The IRS (1997) estimates that about 25.8 percent of EITC recipients in 1994 were actually ineligible for the tax credit. Second, EITC eligibles who receive the credit tend to be the families that would most benefit from the credit. For example, a family that expects to receive a $50 credit is less likely to apply than a family that expects to receive a $3,000 credit. Thus, we expect the eligible population to receive lower average credit amounts than the recipient population.

Determining EITC Eligibility in the SIPP

Beginning in 1995, eligible claimants fall into three groups: childless adults between the ages of 25 and 64, individuals with one qualifying child, and individuals with two or more qualifying children. To claim a qualifying child three tests must be satisfied: (1) a relationship test—-the child must be the son, daughter, step-child, grandchild, foster child, or adopted child of the claimant, (2) an age test—the child must be under 19, or under 24 and a full-time student for at least 5 months, or permanently and totally disabled, and (3) a residency test—the child must live with the claimant for more than half the year.

All claimants must also meet certain income requirements that are discussed in greater detail in the text (see Table 6). In short, claimants must have earned income and adjusted gross income (AGI) that fall within a certain range. AGI is used in cases in which it exceeds earned income and falls in the phase out range of the credit. Disqualified income—including taxable and non-taxable interest income, dividend income, net rent and royalty income, capital gain net income, and net passive income—must not exceed $2,200.

To determine whether individuals would have been eligible for the EITC under the 1996 rules using the SIPP data, we first calculate income and number of qualified children for all individuals. Married couples are assumed to file jointly, while all others file separate returns. Thus, for single individuals we utilize personal income information only, while for married couples we sum each spouses' income data. We count all earnings for the 1990 calendar year, including wage and salary, self-employment, and farm income. The adjusted gross income variable is the sum of earned income, interest income, rents, royalties, net capital gains, unemployment insurance, taxable amounts of Social Security (50 percent of payments to single taxpayers with incomes over $25,000—over $32,000 for married filers), alimony, and 75 percent of pension and retirement income. We include 75 percent of pension and retirement income based on Scholz' (1994) estimate of the amount of pension and retirement income included in AGI.

The number of qualified children is the sum of: 1) the number of own children under 18 years old in each family or subfamily for at least 7 months in 1990; 2) the number of adults in the family who are 18 years old in December 1990 (and not already counted by the above rule); and 3) the number of full-time students in the family between 19 and 23 years old (as of December 1990). A full-time student is one who is enrolled in school full-time in at least 2 waves in 1990. It should be noted that older children under this definition are not necessarily the claimants' own children and are not subject to the residency test. Also, we do not include disabled adults 19 and over in the count of qualified children.

Validation Exercise

As noted previously, our estimates are based on the number of families and individuals eligible for the EITC. Using 1990 data and 1994 EITC rules, we estimate 16 million eligible families and individuals. The IRS (1997) estimates that 19 million tax filers received the EITC in 1994. Our estimate differs from the IRS estimate for the following reasons: 1) noncompliance; 2) nonparticipation; and 3) changes in income and demographics of the population between 1990 and 1994.

A recent study by the IRS estimates that 25.8 percent of the total EITC dollars claimed in 1994 should not have been claimed. This report did not estimate the percent of EITC recipients who were ineligible, but an earlier study (Scholz 1994) found that nearly all of EITC overclaims were made by ineligibles. On the other hand, many families and individuals eligible for the EITC fail to claim the credit. Scholz (1994) estimates that in 1990, only 80 to 86 percent of eligible taxpayers received the credit.

If we assume that 25.8 percent of 1994 recipients are ineligible, then the number of eligible recipients in 1994 would have been 14.1 million (74.2 percent of 19 million). If we further assume an 80 to 86 percent participation rate, then the number of eligibles in 1994 would have been between 16.3 and 17.6 million (14.1 million divided by 86 and 80 percent, respectively). This is close to our estimate of the number of eligibles using 1990 data and 1994 eligibility rules (16 million). Some difference is not surprising, considering changes in income and demographics between 1990 and 1994, and the need for some simplifying assumptions in the modeling process.


Notes

1. These numbers were calculated using data from the Internal Revenue Service's 1993 Individual Statistics of Income Public Use File.

2. For example, in 1996, a single filer would have been taxed at a rate of 28% on taxable income exceeding $24,000. However, a head of household filer would continue to be taxed at the 15% rate up to a taxable income of $32,150. In 1996, the phase-out of personal exemptions began at $117,950 for single filers, and $147,450 for head of household filers.

3. Estimating the percent of income spent on children in in-tact families and incorporating this information into child support guidelines was a major challenge in guideline development. See Bassi and Barnow (1993) for further discussion of the various approaches taken to this problem.

4. Bassi et al (1990) and Holden and Smock (1991) provide a detailed review of this literature.

5. Weitzman (1985) examines the situation of custodial and noncustodial parents assuming full payment of child support.

6. Duncan and Hoffman (1985) include cases involving the departure of a cohabiting parent.

7. In many states child care and health care costs will not have much effect on the relative well-being of the noncustodial parent and custodial family, since guidelines often add child care and health insurance costs to the basic child support award and allocate them between the parents in proportion to income.

8. This probably overestimates the number of noncustodial fathers who would be eligible, since some of these custodial families receive child support from noncustodial fathers outside the income range covered by the EITC. On the other hand, some eligible noncustodial fathers pay child support to custodial families outside the income range covered by the EITC, but there are probably fewer of these cases.

9. This is consistent with the overall tax treatment of child support. Since the noncustodial parent has already paid taxes on income paid in child support, child support is not taxed to the custodial family.

10. A recent GAO report (1997) states that one of the most common types of EITC noncompliance, according to the U.S. Treasury Department, is that tax filers claim qualifying children who do not, in fact, reside with them for more than half the year.


Tables

Table 1. The Effect of Federal Income Taxes, EITC, and Child Support on the Noncustodial Parent, Custodial Family, and Married Parent Family Assuming $12,000 in Earnings for Each Parent, and One Child

Noncustodial Parent Custodial Parent Married Parent Family

Annual Earnings $12,000      $12,000      $24,000     
- Payroll Taxes $918 $918 $1,836
- federal income tax $818 $150 $1,448
+ EITC 0 $2,096 $173
Income after taxes and EITC $10,265 $13,028 $20,889
Child Support ($2,001) $2,001 NA
Income after taxes, EITC, and Child Support $8,264 $15,029 $20,889
Poverty Threshold $8,163 $10,562 $12,517
Percent of Poverty:
    Before Taxes, EITC, and Child Support 147% 114% 192%
    After Taxes, EITC; Before Child Support 126% 123% 167%
    After Taxes, EITC, and Child Support 101% 142% 167%


Table 2. Poverty Level 1 of Noncustodial Parent, Custodial Family, and Married Parent Family Before and After Taxes 2, EITC, and Child Support 3 Assuming $12,000 in Earnings for Each Parent, and One Child

  
KEY: Noncustodial Parent Custodial Family Married Parent Family

   If No Child Support Paid      If Full Child Support Paid  
Poverty Level
Before Taxes
and Before EITC
  
 
   If No Child Support Paid      If Full Child Support Paid  
Poverty Level
After Taxes
but Before EITC
  
 
   If No Child Support Paid      If Full Child Support Paid  
Poverty Level
After Taxes
and After EITC
  

1. Child support is subtracted from the income of the noncustodial parent and added to the income of the custodial family before calculating the percent of poverty. The percent of poverty is based on preliminary 1996 poverty thresholds from the Social Security Administration.

2. Taxes include 1996 federal income taxes and the worker share of social security and medicare payroll taxes.

3. The child support award is calculated according to the median state child support guideline for states using an income shares guideline based on pre-tax income.


Table 3: Poverty Level of Noncustodial Parent, Custodial Family, and Married Parent Family, Before and After Taxes, EITC, and Child Support, Assuming $12,000 in Earnings for Each Parent, and Two Children

  
KEY: Noncustodial Parent Custodial Family Married Parent Family

   If No Child Support Paid      If Full Child Support Paid  
Poverty Level
Before Taxes
and Before EITC
  
 
   If No Child Support Paid      If Full Child Support Paid  
Poverty Level
After Taxes
but Before EITC
  
 
   If No Child Support Paid      If Full Child Support Paid  
Poverty Level
After Taxes
and After EITC
  

1. Child support is subtracted from the income of the noncustodial parent and added to the income of the custodial family before calculating the percent of poverty. The percent of poverty is based on preliminary 1996 poverty thresholds from the Social Security Administration.

2. Taxes include 1996 federal income taxes and the worker share of social security and medicare payroll taxes.

3. The child support award is calculated according to the median state child support guideline for states using an income shares guideline based on pre-tax income.


Table 4: Poverty Level of Noncustodial Parent, Custodial Family, and Married Parent Family, After Federal Income and Payroll Taxes and Child Support, For Different Combinations of Noncustodial and Custodial Parent Earnings, and One Child

  
KEY:Noncustodial Parent Custodial Family Married Parent Family

Earnings Noncustodial $8,000 $12,000 $15,000 $18,000

Custodial



$9,000


$12,000
 
Earnings Noncustodial $8,000 $12,000 $15,000 $18,000

Custodial



$15,000

$18,000


Table 5: Poverty Level of Noncustodial Parent, Custodial Family, and Married Parent Family, After Federal Income and Payroll Taxes and Child Support, For Different Combinations of Noncustodial and Custodial Parent Earnings, and Two Children

  
KEY:Noncustodial Parent Custodial Family Married Parent Family

Earnings Noncustodial $8,000 $12,000 $15,000 $18,000

Custodial



$9,000


$12,000
Earnings Noncustodial $8,000 $12,000 $15,000 $18,000

Custodial



$15,000

$18,000

Table 6: 1996 Earned Income Credit Parameters

Number of Qualifying Children Credit Rate
(Percent)
Minimum Earnings
for Maximum Credit
Maximum Credit Phase out Rate
(Percent)
Phase out Range
  Begins             Ends        

No Children 7.65    $4,220    $323    7.65    $5,280    $9,500   
1 Child 34    $6,330    $2,152    15.98    $11,610    $25,078   
2+ Children 40    $8,890    $3,556    21.06    $11,610    $28,495   
                                      

Table 7: Characteristics of 1996 EITC Eligibles

   Overall   Qualifying Child Credit   Childless Credit  

Number of Eligibles 16.3 Million 11.6 Million 4.7 Million
Average EITC Benefit $1,180 $1,596 $167
Total Cost of the Credit $18.8 B $18.0 B $0.8B
 
Percent of EITC Benefits                  
By Poverty Status                  
Less Than 100% 33%        33%        40%       
100-200% 50%    50%    31%   
More Than 200% 17%    17%    28%   
 
Percent of EITC Recipients,                  
By Poverty Status                    
Less Than 100% 29%    26%    37%   
100-200% 47%    52%    35%   
More Than 200% 23%    21%    28%   

Note:These families would have been eligible for the EITC in 1990 if the 1996 EITC rules had been in effect at that time.


Table 8: Characteristics of EITC Eligible Custodial Mothers and Noncustodial Fathers

   Custodial Mothers   Noncustodial Fathers  

      With a
Qualifying
Child
  
Eligible for
Childless
Credit
  

        
Number of Eligibles 4.7 Million 1.3 Million 0.8 Million
Average EITC Benefit $1,644 $1,602 $185
Total Cost of the Credit $7.2B $1.1B $.1B
           
Percent of EITC Received
by Poverty Status
  
        
Less than 100% of poverty 29% 40% 39%
100-200% of poverty 49% 54% 33%
More than 200% of poverty 22% 6% 20%
           
Child Support Receipt/Payment           
Percent who received/paid 45% 27% 28%
Average amount received/paid $3,153 $2,284 $1,625
As a % of income 14% 23%

Note: These individuals would have been eligible for the EITC in 1990 if the 1996 EITC rules had been in effect at that time.

Source:1990 SIPP.


Table 9: Costs and Benefits of Increasing the EITC to Low-Income Noncustodial Fathers Who Pay Child Support

Policy Options  

Costs and Benefits   1 2 3

           
Number Who Would Benefit 747,000 349,000 743,000
Mean Increase in EITC Benefit $1,286 $1,286 $706
% Increase in EITC Cost 5.1% 2.4% 2.8%
           
Percent of Additional EITC Benefits Received by Poverty Status           
           
Less than 100% of poverty 2% 2% 3%
100-200% of poverty 47% 47% 48%
More than 200% of poverty 51% 51% 49%
           
Average Child Support Paid $3,454 $3,454 $3,438
as a % of Income 20% 20% 20%


Table 10: Poverty Level of Noncustodial Parent, Custodial Family, and Married Parent Family, After Taxes, EITC, and Child Support, Under Different EITC Scenarios, Assuming $12,000 in Earnings for Each Parent

  
KEY: Noncustodial Parent Custodial Family Married Parent Family

Scenario One Child      Two Children  
 

Current Law

Scenario One Child      Two Children  
 
Policy Option 1
Financing Method 1
Scenario One Child      Two Children  
 
Policy Option 3
Financing Method 2
  

1. Child support is subtracted from the income of the noncustodial parent and added to the income of the custodial family before calculating the percent of poverty. The percent of poverty is based on preliminary 1996 poverty thresholds from the Social Security Administration.

2. Taxes include 1996 federal income taxes and the worker share of social security and medicare payroll taxes.

3. The child support award is calculated according to the median state child support guideline for states using an income shares guideline based on pre-tax income.



Topics/Tags: | Children and Youth | Economy/Taxes | Poverty, Assets and Safety Net


Usage and reprints: Most publications may be downloaded free of charge from the web site and may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact publicaffairs@urban.org.

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.

Disclaimer: The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute.

Email this Page