TAXATION OF THE FAMILY

Statement before the Committee on Ways and Means

United States House of Representatives

April 15, 1997

C. Eugene Steuerle

Senior Fellow

Any opinions expressed herein are solely the author's and should not be attributed to The Urban Institute, its officers or funders.

THE URBAN INSTITUTE

2100 M Street, N.W.

Washington, DC 20037

The family is the primary social structure in the United States for nurturing, raising and paying for the needs of children. Support for families, however, has long been an expressed policy goal of both major U.S. political parties. Whether U.S. tax policy -- as opposed to expenditure policy -- should be designed specifically to benefit families is an issue of legitimate debate. In a recent article with Professor Michael McIntyre of Wright State University⁽¹⁾, we subscribed to the traditional view that a personal tax system should be designed primarily to distribute tax burdens in a way that is fair to all individuals, irrespective of their family circumstances. At the same time we concluded that a tax system cannot be fair to individuals unless it takes into account the differences in ability to pay that result from the way that resources are shared within families of different sizes and types.

When it comes to tax reform, "family" issues are often among the last to be considered. In practice, however, these issues often dominate the revenue adjustments that might be required in moving to any different type of system. Having served as the Economic Coordinator for the Treasury Department's 1984 to 1986 tax reform effort, I can warn anyone trying to redesign the tax system that "family" issues at times will drive the process if for no other reason than that so much money is at stake. Congress' recent debate over a child credit demonstrates just how expensive changes here can be. One reason is that decisions over how to treat children or spouses in the tax Code typically involve millions, tens of millions, or even hundreds of millions of people. A change of $500 for 50 million taxpayers, for instance, might require that $25 billion in tax liabilities be shifted annually. By way of contrast, most other reform issues involve far smaller numbers of taxpayers.

Many provisions of the current federal income tax are specifically designed to take into account the economic circumstances of the family. Examples are included in Table 1:

__________________________________________________________________________

Table 1

SOME FAMILY-SENSITIVE TAX ISSUES

• the dependency exemption;
• a child credit (proposed by both major political parties, but not enacted);
• the earned income tax credit;
• the child and dependent care credit;
• the "kiddie" tax;
• the standard deduction and tax-free levels of income that vary in amount for different types of households;
• the special rate schedule for heads of households;
• marital income splitting and the rate schedule for single taxpayers; and
• the deduction for alimony payments and the nondeductibility of child support payments.

____________________________________________________________________________

Obviously, all tax issues affect the family in some manner or another. For instance, the home mortgage interest deduction and the charitable contributions deduction affect families in different ways. However, I will confine my discussion today mainly to those issues where adjustments in tax burden are made according to the size of the family -- in particular, the presence of children.

One major source of complication must be admitted up front. Current tax law includes several measures designed to benefit low-income families. Some of these measures are defended on tax policy grounds, whereas others are defended on spending policy grounds. A major objective of family taxation reform -- indeed, one that has become unavoidable -- should be to coordinate the tax measures that are designed to benefit low-income families with children with the various direct expenditure programs targeted at such families. Indeed, as I will demonstrate, tax administration often requires this coordination whether we desire it or not.

a. Dependency Exemption

The dependency exemption is a major mechanism for adjusting tax burdens for the costs of supporting children and is the only mechanism that provides tax benefits to all middle-income families with dependent children. The dependency exemption for 1996 was $2,550 per dependent child. Since 1987, it has been phased out for high-income taxpayers. For tax year 1994, taxpayers claimed a total of approximately 70 million dependency exemptions.

Few changes in federal income tax laws over the past four decades have had as far-reaching effects on the distribution of federal tax burdens as the shift in the relative tax burdens from taxpayers without dependent children to taxpayers with dependent children. The increase in relative tax burdens has been particularly marked for middle-income taxpayers with children.

The increase in relative burdens on families with dependent children did not occur because policymakers, after careful study, concluded that parents with dependent children were being taxed too lightly. Instead, it happened primarily because the chief mechanism for granting tax relief to families with dependent children -- the dependency exemption -- was not adjusted sufficiently to keep up with economic growth.

Over the 48-year period from 1948 to 1996, the dependency exemption has grown from $600 to $2,550 -- slightly more than a four-fold increase. During that same period, per capita personal income has grown from $1,425 to $23,882, which is more than a sixteen-fold increase. As a consequence of economic growth, the dependency exemption fell from about 42 percent of per capita personal income in 1948 to less than 11 percent by 1996.

The dependency exemption for 1996 would need to be set at approximately $10,000 for it to represent the same percentage of per capita income as it represented in 1948 (Figure 1). Simply to adjust the dependency exemption for post-1948 inflation would require that it be increased to nearly $4,000. If converted to a credit that offset taxes, the exemption would need to equal $1,500 or more per child to reduce taxes for the same proportion of income as in 1948.

This decline in the dependency exemption, along with increases in Social Security taxes, has increased the tax burden of families with children relative to almost all other taxpayers and is one of the major reasons today for the consideration of a child credit.

b. Tax Credit for Dependent Children (Child Credit)

A child dependency credit, generally referred to as a child credit, is a possible alternative mechanism for delivering tax relief to parents with dependent children. The credit might be a fixed amount per dependent child, or the amount of the credit might vary with family size. It could be fixed in amount at all income levels; some would phase out at middle- or high-income levels, although phase outs by their very nature involve implicit rather than explicit tax rates. Although both political parties have proposed child credits, most major reform proposals do not deal with this issue.

Significant simplification gains are possible if tax relief to families with dependent children is provided through one mechanism that integrates the benefits of current law and any new benefits that policymakers are prepared to give to families with dependent children. Under current law, relief is now targeted at families with children through the dependency exemption and, for low-income families, through the earned income tax credit (EITC). Adding a third relief mechanism with a different set of eligibility rules appears needlessly complex. Nonetheless, rolling two and perhaps all three mechanisms into a single mechanism would require some changes in policy -- for example, a uniform definition of "dependent" would probably be required. Such changes may create some additional winners and losers in order to achieve gains in administrative economy.

In tax theory, there is no strong case in favor of a credit over a deduction or vice-versa. Indeed, as a technical matter, for families of one size but different incomes, it is possible to develop a credit-based system that would replicate exactly an exemption-based system. Assuming the continuation of the EITC and the commitment of the nation to provide minimal levels of support to many low-income individuals, the use of a unified credit would seem to be the preferred approach. In addition, a credit mechanism that combined the benefits of the EITC and the dependency exemption could be coordinated better with various rules for phasing out welfare benefits than is possible under current law.

Finally, I have also suggested that Congress may have a unique opportunity today to link some unified child credit with a requirement that the credit is only available to families that purchase health insurance for their children. In one combined effort, Congress could partiality reverse the historical trend toward increasing the relative tax burden placed on families with children, reduce dramatically the lack of health insurance among children (and among some adults as well), and reduce some of the high implicit tax rates imposed on those who decide not to take welfare or who move off of welfare.

c. Earned Income Tax Credit

Current law provides low-income workers with a refundable tax credit -- that is, besides reducing the tax liability for low-income families, the government sends a check to the taxpayer for any amount by which the allowable credit exceeds that taxpayer's liability for taxes payable on his or her income tax return. This earned income tax credit (EITC) provides significant benefits to low-income families with dependent children and more limited relief to other low-income individuals. Taxpayers with income over specified income thresholds are not eligible for the EITC.

The EITC began as a limited program in 1975 during the Ford administration and has been expanded several times since then, with large increases enacted in 1986, 1990, and 1993 and some modest adjustments in 1996. The 1993 additions were only scheduled to become fully effective in 1996. Most of the credits represent amounts refunded to households.

Historically, the EITC has been promoted as a useful mechanism for lowering income taxes and offsetting FICA (Social Security) taxes for low-income individuals with dependent children; for some it also offset the work disincentives associated with welfare. Both of these goals continue to be invoked to justify the EITC. Today, the EITC probably should be considered primarily as an extension of a combined welfare/tax system. That is, it has important tax and welfare features that to some extent are inseparable.

Proponents of many types of tax reform have many difficult issues to face with the EITC. For example, it is impossible to remove most low-income families from filing returns as long as the EITC is designed in its current fashion. For those who favor many types of consumption taxes, the EITC also affects the simplification gains that they hope to achieve by eliminating requirements to measure capital income for tax purposes. The EITC is necessarily income-based, unless high-income individuals with low wages or low levels of consumption are to be made eligible for the EITC. To administer the EITC, therefore, the tax authorities must obtain substantial information about the capital income of prospective recipients of the EITC. A similar constraint applies to welfare authorities administering Food Stamps, Supplemental Security Income, and other programs. The problem is not solved by making the issue of relief for low-income workers a problem for the welfare system. Those systems contain implicit income taxes that affect millions. Meanwhile, businesses, banks, and other institutions would still need to perform income reporting even if the main body of the income tax were converted to a consumption tax.

Revision of the EITC is likely to be a topic on the public agenda for some time, whatever the political fate of the major reform plans. The EITC has received political support from many sources, including, at various times, from the leadership of the two major political parties. In my view, it represents an intermediate step as the nation searches for a way to convert welfare into work support. It has been favored both because of the work incentives that it provides compared to welfare, as well as the relief that it delivers to low-income families. Some supporters have seen it as a politically viable alternative to an increase in the minimum wage. The EITC, however, has also received criticisms from a range of sources, partly because of problems with its implementation and partly because it is not well integrated with income-tested welfare programs. Congress and the IRS have attempted to deal with the problem of ineligible participants receiving the credit by reforming eligibility criteria and by checking more closely with taxpayers over the existence of dependents. Error rates, however, remain very high.

An additional problem in the EITC remains to be addressed: the ability of taxpayers to overdeclare income to receive higher credit amounts. This problem, which I have labeled the "superterranean economy" (as opposed to the underreporting of income in the subterrranean economy), does not require cheating. Two neighbors could baby-sit for each other and generate significant EITCs as a consequence.

Coordination of Tax Provisions with Implicit Taxes Embedded in Welfare Programs. I recognize that your focus today is on tax reform, while a focus on family issues keeps pulling us toward discussion of transfer programs as well. For low and moderate income individuals, however, tax and transfer issues simply can no longer be separated. An important objective of public policy, whether characterized as tax policy or welfare policy, should be to substantially reduce the high marginal "tax" rates that low income individuals typically face when they attempt to enter the workforce or the effective tax rates that low income workers face simply by choosing never to go on welfare. A reduction in those rates presumably would discourage long-term dependence on welfare. It would also reduce the extent to which low-income workers perceive that they are being treated unfairly.

Figure 2 shows the combined tax rates derived from tax and welfare programs just before the enactment of welfare reform -- although it is doubtful that these rates have changed much since then. In effect, welfare recipients who worked faced combined tax rates of 70 percent not just when they went to work at minimum wage but all the way up toward three times the minimum wage (the effective marginal rate on additional work is often even higher than this "average" rate on all earnings). At one to three times minimum wage for a full-time worker, few individuals receive much in what is commonly thought of as welfare: AFDC or its replacement, TANF (Temporary Assistance to Needy Families). The high tax rates at those income levels derive from federal income tax, phase out of the EITC, state income tax, phase out of housing benefits for those who receive them, phase out of Food Stamps, and phase out of eligibility for Medicaid. All phase outs, remember, basically take away benefits as income increases. Avoidance of such poverty traps should be an important, long-term, objective of any major tax reform.

In addition to the large tax rates on work, marriage penalties are enormous for low-income individuals. For a typical welfare recipient who married someone with a modest paying job, their combined income would fall by an additional 30 percent or so just from marriage alone. Many marriage penalties are caused by welfare, but some are in the tax code itself due to the earned income tax credit and the standard deduction. Thus, another potential advantage of a unified approach to tax and welfare issues is the opportunity provided for reducing marriage penalties. Although, as discussed later, there are also marriage penalties for higher income individuals due to the rate structure, but these are smaller relative to income than those faced by low income individuals.

Let me return briefly to how a child credit provides a means of linking together these concerns between the welfare system and the tax system. Just as a welfare payment operates as a refundable tax credit that is phased out as income increases, so also a child credit could be designed to be there when the welfare payment was no longer available. Once the credit is fully phased in, it can be allowed to remain constant throughout the low- and middle-income ranges, thereby avoiding the implicit taxes that result under current law from the phase-out of the EITC or welfare credits.

In effect, a child credit can be explicitly designed to reduce, although not eliminate, some of the poverty traps and marriage penalties faced by low income individuals. As far as I can tell, none of the major tax reform proposals on the national agenda attempt to address the poverty traps created by the interplay of tax and welfare policies. Several of them attempt to replicate the distribution of taxes at lower income levels and simply to leave these issues to another time and place.

d. Child-Care Credit

Parents with one or more children under age 13 may claim a tax credit under current law for a portion of the expenses for child care and household services that they incur in pursuing gainful employment outside the home. The allowable credit is a percentage (30 percent at low-income levels, phased down to 20 percent) of qualifying expenses. Qualifying expenses are capped at $2,400 (one qualifying dependent) or at $4,800 (two or more qualifying dependents). In the case of a two-job married couple, the expenses eligible for the credit generally cannot exceed the income of the lower-earner spouse. Taxpayers claiming the credit must provide the Internal Revenue Service with the name, address, and taxpayer identification number of their provider.

A deduction for child-care expenses was introduced in 1954, during the Eisenhower administration, primarily as a mechanism for encouraging mothers on welfare to work outside the home. The deduction was capped at $600 and was phased out at rather low income levels. The allowance has been expanded several times and was converted into a credit in 1976. The child-care credit was claimed in 1995 by just over 6 million taxpayers for total credits of under $3 billion. For a taxpayer with adjusted gross income of $10,000 or less and two qualifying dependents, the maximum credit is $1,440. The maximum credit is $960 for parents with income of $30,000 or more.

The case for repeal on efficiency grounds is at best mixed. An initial and continuing purpose of a child-care allowance has been to mitigate the tax and welfare disincentives that some parents face in taking a job in the labor market. The efficiency problem arises if the tax system tries to be neutral between child care provided in the home and child care provided outside of the home. The current credit generally favors child care outside of the home for those with low and middle incomes, but favors child care in the home for taxpayers with above average incomes. Eliminating any adjustment for child care clearly would favor child care in the home, as can be seen most readily by examining the circumstances of a single parent who must obtain child care in order to work. Maintaining an adjustment, on the other hand, would cause some modest increase in the tax rate, which could have some efficiency costs.

The case for a child-care allowance on fairness grounds also is mixed. One argument for a child-care allowance is that it constitutes a legitimate cost of earning income and ought to be deductible in a tax system seeking to measure net income (or net consumption). Those arguing that child-care expenses constitute a business cost can show that the costs of child care are closely analogous to certain expenses, such as the costs of travel away from home, that are deductible as a cost of earning income. On the other hand, those costs are also analogous to certain other expenses, such as the cost of most types of personal clothing, that are not deductible, notwithstanding a close relationship to business. Because child-care costs arise from the quintessentially personal decision to have and raise children, a case for the deduction on business-expense grounds can never be conclusively made.

As a practical matter, I believe that some adjustment is appropriate but needs to be limited and kept simple. Nonetheless, any reform proposal that attempts to eliminate filing requirements cannot maintain a child care credit or deduction unless these could be channeled directly through employers.

e. Kiddie Tax

Under current law, as amended in 1986, children under the age of 14 are taxable on their unearned income at the marginal tax rate of their parents. This rule is popularly, if inexactly, referred to as the "kiddie tax." Its point is to prevent parents from avoiding the bite of the graduated rate structure by shifting investment income to their children. Its initial purpose was to simplify tax planning costs. Its adoption, indeed, did reduce the tax planning benefits obtaining from establishing certain family trusts, thereby reducing the complexity for the taxpayer and the tax authorities that is associated with such tax planning. Earned income -- e.g., income that children earn from babysitting or delivering newspapers -- is not subject to the kiddie tax rule. Nonetheless, in 1986 Congress went much further than the initial "kiddie tax" goals when it dramatically reduced or eliminated the personal exemption for children with earnings who were also claimed by their parents. This created much additional filing complexity and a significant increase in children required to file. Simplification requires a restoration of something like an additional personal exemption for children, even though on strict equity grounds some children would thereby generate more personal exemption than others.

f. The Standard Deduction and Tax Free Levels of Income

The tax-free level under current law is determined by two mechanisms: the taxpayer exemption and the standard deduction. For 1996, the taxpayer exemption was set at $2,550. This is the same amount as the dependency deduction. These personal exemptions were set at $2,000 after the phase-in of the 1986 tax act (1989) and have been adjusted upwards for inflation since then.

Each type of filing unit has its own standard deduction level. For married couples in 1996 it was $6,700 for a per capita standard deduction of $3,350. Heads of household received a standard deduction of $5,900, while single individuals could claim a standard deduction of $4,000. Relative to being single, the standard deduction creates a modest marriage penalty for many moderate income couples typically amounting to $195 in extra tax liability.

Personal exemptions, on the other hand, are of equal size for all persons. Together with the standard deduction they provide for a tax-free level of income of $6,550 for a single individual, $8,450 for a head of household, and $5,900 for each member of a couple (Table 2). Excluding the earned income tax credit (EITC), most reform proposals would increase this tax exempt amount. Only flat and retail sales tax proposals usually remove marriage penalties from this source, although there is no reason that other reforms could not also achieve that goal.

g. The Head of Household Schedule

The head-of-household schedule was introduced into the tax code in 1951. Its purpose was to extend to one-parent families some portion of the tax benefits that two-parent families received under the marital income splitting regime adopted nationally in 1948. Under that regime, marital partners were allowed to report one-half of the total income of their marital partnership on the same rate schedule used by single individuals. In contrast to the head of household schedule, the benefits of marital income splitting were available to all marital couples, whether or not they had dependent children.

The purpose of the head-of-household schedule is to take account of the differences in ability to pay of heads of households relative to equal-income single individuals due to the difference in their support obligations. In effect, the head of a one-parent family is allowed to split income with a dependent child, with the child's portion of the parent's income being taxed at a low or zero rate. The head-of-household schedule operates like a dependency exemption that increases in value with increases n the total income level of the one-parent family.

The special rate schedule for one-parent families creates the potential for marriage penalty because a husband and wife with children could reduce their taxes under current law by getting a divorce, using the deduction for alimony to equalize their individual incomes, and then having one former spouse file as a head of household and the other spouse file as a single person. The former spouses cannot both file as a head of household under current law and still live together, because a head of household is defined as a person providing more than half of the cost of maintaining the household. It does not appear that significant numbers of married couples have availed themselves of this tax-avoidance opportunity.

Heads of household bore significantly higher tax burdens because of the decline in dependent exemptions noted above. If child allowances were raised significantly, this would reduce the need for a separate head of household rate schedule.

Source: Michael J. McIntyre and C. Eugene Steuerle, Federal Tax Reform: A Family Perspective (1996, p. 44).

h. Marital Income Splitting

The modern history of the current federal system of marital taxation begins in 1948, when Congress adopted marital income splitting as a conscious federal policy. Before the 1948 reform, federal family taxation policy was in disarray.

In a tax system that provides for full marital income splitting, each spouse is taxable as an individual on one-half of the total income of their marital partnership. Such a system is not designed primarily to benefit dependent children. It is available, after all, to childless couples and to couples with adult children no longer dependent on their parents. Its purpose is to tax each spouse on that share of the total income of their marital partnership that is used to enhance their material well-being. It can be viewed as implementing the traditional income tax policy goal of relating the burdens of taxation to the consumption and net savings of individual taxpayers.

In 1969, Congress adopted a special tax rate for single individuals that guaranteed that they would pay no more than 120 percent of the tax imposed on marital partners having the same aggregate income. This 120-percent rule reflected a political compromise between those who contended that equal-income marital couples should bear equal taxes and those who contended that individuals with equal income or equal earnings should pay equal taxes notwithstanding differences in their marital status. The revenue cost of introducing the "singles" rate schedule was modest -- on the order of $200 million per year in forgone revenue. Despite the low cost, the implications of this change for federal tax policy were very large, for reasons explained below.

Under the system adopted in 1969, marital partners became taxable on their aggregate incomes as a unit, under a rate schedule with brackets exactly twice as wide as the brackets under the rate schedule of prior law. The tax brackets on the marital unit schedule, however, were less than twice as wide as the brackets on the newly created schedule for single persons. The effect was that two marital partners having approximately equal separate incomes would pay less in tax if they were allowed to file separate tax returns and to compute their separate tax liabilities on the new singles schedule. The only way to do so, however, was to terminate their marriage. The tax savings that marital partners could obtain from getting a divorce and filing separately came to be called a "tax on marriage" or a "marriage penalty."

Congress has adopted legislation from time to time to reduce the marriage penalties created by the 1969 act. Other legislation, unfortunately, has increased those penalties. Marriage penalties were reduced sharply under the 1986 tax act, due to the flattening of the rate structure and the introduction of fuller income splitting at middle-income levels. Marriage penalties were increased significantly by the way that the 1993 tax act increased tax rates for high-income individuals. No changes have been made in the basic system of multiple graduated rate schedules introduced in 1969, which necessarily produces marriage penalties. Plans that attempt to replicate the existing distribution of tax burdens, such as the USA plan and Gephardt 10-percent tax, tend to continue that basic structure.

A perfectly flat tax would eliminate all marriage penalties created by the graduated structure. This approach, combined with the equal per capita standard deductions provided to single and married persons, would eliminate almost all marriage penalties created by the rate structure, although not -- as mentioned above -- the very large penalties due to welfare and EITC type of provisions. To eliminate marriage penalties, one also needs to eliminate phase-outs such as the phase out of itemized deductions and the personal exemption phase out.

i. Alimony and Child Support Payments

Under current law, alimony is deductible to the payor and taxable to the recipient. The effect of this arrangement is to extend some degree of income splitting to formerly married individuals. Thus, the treatment of current law is consistent with the income splitting approach to family taxation.

In the typical case, alimony flows from the higher-earner taxpayer to the lower-earner taxpayer. In a tax system having graduated rates, therefore, taxing the recipient of alimony rather than the payor results in a net reduction in the aggregate tax burdens of the two former spouses. If the tax savings to the payor and the tax detriment to the recipient are properly taken into account in setting the level of the alimony payments, the alimony recipient should obtain a net benefit from having been made taxable on the alimony payments. That is, the recipient would receive an additional alimony payment sufficient to pay the tax and to give that individual some fair share of the resulting tax savings. Divorce settlements that provide for the payment of alimony are typically structured so that they deflect some or all of the tax savings from the alimony deduction to the alimony recipient.

The proper tax policy treatment of child-support payments is unclear. Those who hold that the earner is the proper taxpayer on earned income presumably would oppose the deduction of support payments. The earner rule, however, is inconsistent with marital income splitting -- an approach endorsed under current law and under several reform proposals. If an income splitting approach is carried over to children, then child-support payments would be deductible to the payor and taxable to the child, not to the custodial parent. It certainly would be an odd result, however, to allow income splitting between separated parents and their children and to not allow it within fully intact families.

As discussed above, the dependency exemption can be understood as a mechanism for allowing limited income splitting with children. If the dependency exemption, or a child credit, is generous, then the issue of who to tax on support payments has reduced importance, because the parent taxable on the support payments presumably would be the one who would be allowed to claim the dependency exemption or credit.

Despite all of these arguments, perhaps the simplest system administratively, and the one with the fewest enforcement problems, is to tax income to the earner and to grant exemptions and credits primarily upon the basis of with whom the child lives most of the year.

In a tax system with graduated tax rates, a rule that taxed child-support payments to the recipient parent and made them deductible by the payor parent typically would result in lower aggregate taxes on those parents, assuming that the payments flow from the higher-bracket taxpayer to the lower-bracket taxpayer. The point is similar to one that can be made with respect to alimony payments. Both parents would be better off under a deduction rule as long as some mechanism was in place that would require them to share fairly the net tax savings. Even in a single-rate system, such as a flat tax, divorced or separated parents would obtain a net benefit from the deduction rule whenever the recipient parent's income otherwise would have been below the tax-exempt level. For simplification purposes, however, most flat and consumption-based taxes would assume that the flat rate structure eliminated most concerns over who paid tax and would rely upon withholding of the tax at the source of payment, such as the employer or business.

Conclusion

My testimony today has dealt with the many family-sensitive provisions that must be dealt with in any major tax reform effort. Several of these family related issues affect very large numbers of taxpayers and involve significant amounts of revenues. While certain types of reform efforts technically may move beyond an income tax, they often cannot bypass these family issues. Thus, earned income tax credits and welfare programs have income phase outs that operate like income taxes even in the presence of a consumption tax. In a consumption tax or value-added tax or retail sales tax that did not allow for a child care deduction, a decision would still have to be made as to whether child care expenses were to be taxable as consumption services. Some divorce settlements would have to be renegotiated under major tax reform, especially when they were based on the allocation of tax benefits under current law.

Of all the issues I have raised, perhaps the largest and most important are those that relate to the ways that any tax system adjusts for the presence of children through child credits, dependent exemptions, and the earned income tax credit. Recent bipartisan support for child credits and the push for tax reform create a unique opportunity to lessen the increasing reliance of our tax system on families with children, to expand significantly health insurance for children, and, at the same time, to reduce the extraordinarily high tax rates and marriage penalties on those low-income individuals who decide to work rather than rely on welfare.

1. See Michael J. McIntyre and C. Eugene Steuerle, Federal Tax Reform: A Family Perspective, Washington, D.C.: The Finance Project, July, 1996. Most of this testimony is taken from that article. Some additional material is taken from Linda Giannarelli and C. Eugene Steuerle..." Proceedings of the National Tax Association... and from this author's Economic Perspectives columns published by Tax Notes magazine.