Horizontal equity and family tax treatment: the orphan child of tax policy.

INTRODUCTION

There is a good chance that the "Brady Bunch" would not pay income taxes even though they have the wherewithal to hire a housekeeper. The exempt level for a married couple with six children who itemize deductions at a rate of 22 percent of income was approximately $90,000

The contours of our tax system with respect to families of different sizes have changed dramatically in the past 15 years. Child credits and the larger earned income credit (EIC) have significantly reduced taxes, or magnified negative tax rates, for lower and moderate income families with children. The growing Alternative Minimum Tax (AMT) has, at the same time, affected families with children in a way perhaps not originally intended. It will do so to a greater extent, absent change, as time goes on. These changes have taken place without, apparently, appealing to standards of horizontal equity--hence, our characterization of this issue as an orphan child. Is our resulting tax system reasonably equitable across family types? And if not, what structural changes would be needed to make it more equitable? Those questions are the object of this paper.

HORIZONTAL EQUITY ACROSS FAMILIES

Horizontal equity suggests that equals before tax should be equals after tax, or, alternatively, that equals should have the same effective tax rates.

One way of measuring equals is through utility analysis. Such an approach is not possible to implement, but it may suggest that when people choose to have children, their costs are compensated by the increase in utility from doing so through the consumption or investment value of their children. In the case of viewing children as consumption, no exemptions, credits or other tax adjustments should be allowed. In the case of viewing children as an investment, income from support by children in old age should be taxed, but a recovery of costs should be allowed. One can extend this argument to the decision to marry as well, although the utility measure is complicated by competing models of marriage, including bargaining models.

Aside from the difficultly in implementing a utility-based measure, such a measure suffers from another problem. Treating children as consumption does not place any weight on the utility of the children. Even from the point of view of the adults, the utility of having children is uncertain. One may have a dutiful child or a "bad seed." Moreover, couples do not have complete control over having children or over marriage and divorce.

We consider instead an ability-to-pay approach. In this approach, families that have the same standard of living should pay the same effective tax rate. Determining equality must adjust for at least two factors. First, children may not require a different expenditure from adults. For example, they may need less food. Second, there are public goods aspects of family consumption as well as economies of scale in home production. For example, additional family members can share kitchens and other common living areas, the family car, telephone service, and entertainment devices.

Such effects have long been recognized in the formulation of equivalence scales for measuring the poverty lines. We adopt such an approach in our analysis. In particular, we use an equivalence index of the form: [(A + PK).sup.e], where A is the number of adults, K is the number of children, P is the ratio of expenditures on a child to that of an adult, and e < 1 is the measure of public goods or scale economies in the household.

For a single adult, [A.sup.e] = 1. If e = 0, all families with the same income should be taxed at the same rate, as all goods are public. If P = 1 and e = 1, a family of four needs four times the income of a family of one, the two families should face the same tax rate, and the exemptions and rate brackets should be four times as large for a family of four as for a family of one. This latter case suggests full income splitting among all family members. The right answer lies in between these measures. If P = 1 and e = 0.5, a family of four needs twice as much income as a family of one. If P = 0.5 and e = 0.5, a married couple with two children needs 173 percent of the income of a family of one.

Our analysis sets both P and e at 0.7 (a married couple with two children needs 235 percent of the income of a family of one), following the estimates in Citro and Michael (1995). At the end of the paper we consider sensitivity analysis.

IMPORTANT FEATURES OF THE TAX STRUCTURE

We cannot consider all of the many features of the tax code that determine tax liability, but rather focus on the main structural features that are relevant to differential treatment of families. These include personal exemptions; standard or itemized deductions; the rate structures for single, head-of-household, and joint returns; the EIC; the child credit; and the AMT. In later sections we also consider the dependent care credit and the benefit of no tax on the additional home production of a non-working spouse. We do not consider the flat rate taxes on capital gains and dividends since rates at the higher income levels where these forms of income are important converge. All calculations are for 2005.

PHILOSOPHIES THAT DRIVE THE DEVELOPMENT OF TAX FEATURES

Potential inequities across family types derive from the objectives of vertical equity, or a progressive tax system, a welfare philosophy, marriage incentives, and other behavioral incentives.

A flat rate tax would tax all families of all incomes at the same rate. Some differences across families are accounted for in basic features of the income tax, which allow personal exemptions and standard deductions that vary across family types and sizes, as well as different rate structures for single, head-of-household and joint returns. Other aspects of the tax code that are driven by vertical equity objectives are the phaseout of personal exemptions, itemized deductions, earned income credits, child credits, dependent care credits, and the AMT, which applies only at higher income levels.

The second feature of the system appears to derive from philosophies underlying the welfare system. Cash assistance in the U.S. welfare system is directed to only certain individuals reflecting an inability to work: dependent children, and the elderly and disabled. When the EIC, a negative income tax, was adopted, it followed this welfare philosophy by initially allowing only benefits for families with children, although all of the families receiving the credit are working families.

It is not clear why this dependency philosophy was applied rather than allowing benefits reflecting needs. Hoffman and Seidman (2003) cite several reasons, but the initial enactment of the EIC in 1975 was done with virtually no debate or notice. One of the rationales advanced for the EIC was to relieve payroll taxes, suggesting benefits should be provided for all workers. Perhaps a more important initial reason for restricting the EIC to families with children was to encourage these families to exit from the welfare system or offer an alternative to welfare expansion, although this argument might suggest the benefit be limited to single parents. Another possible reason was to avoid providing benefits for dependents whose family incomes may be large, students, or for retired individuals with small earnings and income relative to wealth. In 1993, when the credit was increased substantially, a small credit was allowed for families without children, and the targeting issue appears to play some role. These benefits were restricted to an age threshold of 25 and a ceiling of 65. Despite dealing with this targeting issue through age restrictions, the rate remained much smaller for singles and childless couples. The EIC rate was also increased, in 1990 and 1993, for families with two children, relative to one, but no adjustments for additional children were made. Currently, the EIC credit rate is 7.65 percent for families without children, 34 percent for families with one child, and 40 percent for families with two or more children.

The child credit of $1,000 extends through the income scale (although it is phased out for very high incomes), but is also consistent with this dependency philosophy. The political history suggests it may have originated from a desire to help traditional families.

A third philosophy reflects concerns about disincentives in the tax system to marry. Recent tax changes have eliminated the marriage penalty for many taxpayers in the middle of the income distribution. Penalties remain for marriage when one potential spouse already has children through the ability to qualify for the head-of-household schedule. Penalties also remain in the EIC, due to phaseouts, and at higher income levels for singles who marry. These changes, which reduced the marriage penalty, also increased the marriage bonus and increased the differentials in tax between singles and married couples. These rules affect tax burden across families of different types, and do so differentially across the income classes.

Lastly, as noted above, behavioral incentives may have played a role in the development of the EIC, which encourages labor market participation (although it can produce a disincentive for secondary earnings and for hours due to phaseouts), along with many other features of the tax system that create tax preferences.

Although some adjustments have always been made for family size and characteristics, an explicit consideration of horizontal equity across families, beginning with some attempt to measure differing abilities to pay, has not appeared to play a role in the development of the tax system.

ANALYSIS IN THE BASIC CASE

Our analysis considers the effective tax rates for 14 types of families: married couples with zero through six children, singles, and heads of household with one through six children. In order to provide some idea of the frequency of these types of families, Table 1A provides the distribution by income class for families headed by singles, married couples without children, heads of household, and married couples with children, based on data from the 2003 statistics of income (the sample was too small for head-of-household returns at higher income levels).

Our analysis focuses on working-age families, assuming full eligibility for the EIC. Both younger and older heads without children are not eligible for the EIC. Eliminating younger individuals would significantly change the distribution in the lowest income level. According to Weber (1998), 36 percent of returns in 1993 with incomes under $10,000 were filed by dependents. Assuming that these taxpayers are largely single, eliminating these returns would reduce the share of single returns to less than 63 percent (because income levels would have risen over that period, singles accounted for 71 percent of this class in 1993); and increase head-of-household returns to more than 20 percent of the total, married returns with children to more than six percent, and married returns without children to more than ten percent. According to Sailer, Yau, and Rehula (2001), in 1998, 49 percent of returns with incomes below $10,000 were filed by heads less than 25 years old. Assuming that many of these returns were also filed by singles, adjusting for this age difference would also further reduce the share of single returns in the lower income levels for our analysis. Adjusting for the elderly would reduce both single returns and joint returns without children as a share, but would do so through much more of the income scale; Sailer, Yau, and Rehula report that taxpayers over 65 accounted for 12 percent of all returns, 11 percent of returns under $10,000, and 19 percent of returns in the $10,000 to $15,000 range and in the $15,000 to $20,000 range.

Keeping these adjustments in mind, while all types of taxpayers are represented across the income levels, single returns and heads of household are more common in the lower income levels, while married couples dominate higher income levels, with couples without children becoming a larger share as incomes rise. Due to age restrictions for singles and married couples without children, despite the preponderance of single returns in the lowest income levels, head-of-household returns account for 52.6 percent of returns claiming the EIC, while joint returns account for 24.7 percent and singles account for 22.7 percent (Scott, 2006).

Based on numbers of exemptions for dependents, families with children tend to have relatively few children, with the average number less than 1.8, rising with income to slightly over two at the very high income levels. Families with many children are the exception rather than the rule. According to Census data, which reports a slightly larger average size (around two), single-headed families with six children represent less than two percent of the total of these families with children, and families with six children represent less than one percent of all families with children. Families with five, four, three, and two children, and one child represent three percent, six percent, 17 percent, 36 percent, and 38 percent of families with children, respectively (Fields, 2003). For families receiving the earned income credit, 37.5 percent of earned income credits are claimed for one child, 42.5 percent are claimed for two or more children, and 19.9 percent have no qualifying child (Scott, 2006). Despite the tendencies for certain family types to be more common, families of all types are represented in the taxpayer base and large families account for millions of families.

Table 1B provides the family incomes used in the effective tax rate calculations by family type, using the equivalency scale discussed above. Each cell relates to a reference income level of a married couple with no children, and the incomes range from $10,000 to $500,000.

The analysis in the base case considers the rate structure, the personal exemption, the standard deduction or itemized deductions at 22 percent of income (reflecting the average in the tax statistics), the EIC, the child credit, and the AMT. Table 1C shows the effective tax rates using our equivalency scale for 14 types of families. At higher income levels, itemized deductions tend to even out tax rates because they are assumed to be a fixed percentage of income (although this effect is reduced by the action of the AMT). Tax rates affected by the AMT are bolded. The reference incomes are for childless couples. Figures 1A and 1B show these tax rates across the income scale to provide a visual picture of the variation.

[FIGURE 1 OMITTED]

The effective tax rates vary dramatically at lower income levels. At a $10,000 reference income, all effective tax rates are negative, and the rates range from -1.47 percent for a married couple with no children to -39.21 percent for a head-of-household return with two children, a difference of more than a third of income. At a $15,000 reference income, the rates range from zero to a -27.83 percent rate for a head of household with two children, a difference of over a quarter of income. These differences reflect two important factors: the higher EIC rates for families with children, particularly for two children, and the phaseout of the EIC as income rises. In general, the higher incomes required for a married couple to have the same equivalency as a single-headed family lead to greater phaseouts. But the more powerful effect is the much larger credit rates for children, which result in dramatic differences between families with and without children. When we recall that 20 percent of taxpayer returns claiming the credit do not have children, the discrepancy is significant.

At the $25,000 reference income, with the EIC of lesser importance, the variations become smaller, with a rate of 4.67 percent for singles and a -8.55 percent rate for a head of household with six children. The variation is still large--13 percent of income--and remains significant even for more common family types. At these income levels, joint returns without children are favored relative to singles, a reflection of the provisions that ended the marriage penalty in the middle income ranges, but magnified the marriage bonuses. However, a much more generous treatment of families with children remains, reflecting both the EIC and the child credit. The latter causes the most generous treatment for families with many children, especially among married couples.

All tax rates turn positive at the $50,000 reference income and the variation in effective tax rates is much less--around five percent of income. The highest rates are still paid by singles, but the benefit of children begins to diminish as the number of children increases, reflecting in part the effects of the AMT. The rates range from 2.95 percent to 8.32 percent, a range of about 5.5 percent of income, with the lowest rate paid by a married couple with five children. The lowest rates tend to be paid by married couples with three or more children, and by heads of household with three children, the latter pattern reflecting in part the head-of-household rate structure. Itemized deductions, which are a constant percentage of income, tend to moderate differences across family size. (Itemized deduction shares do vary across incomes, and, in particular, are smaller at very high levels of incomes, but tax rates are relatively even in any case at these high income levels. The possible understatement of itemized deductions offsets the effects of not considering lower rates on capital gains and dividends.) In this instance, the AMT tends to lead to more even rates across family types within each head status, although it magnifies differences between joint returns and single or head-of-household returns.

As incomes continue to rise, large families are increasingly taxed more heavily. At the $75,000 reference income, when incomes become quite high, tax rates range from 8.62 percent to 15.07 percent, a difference of slightly over seven percent of income. The lowest rate is paid by a married couple with two children, while the highest rate is paid by a head of household with six children. In this income class, while benefits for children are large enough to lead to slightly lower rates for families with only a few children, tax rates rise for larger families. As we shall see below, this discrepancy is caused in large part by the effects of the AMT whose exemption does not adjust for family size and whose restrictions on itemized deductions reduce the leveling effects of that provision on effective tax rates.

At this reference income, families with many children pay the highest rates because the personal exemption and child credit are not large enough relative to income to adjust for ability to pay. This effect is increased by the AMT. Tax rates without the AMT would be relatively even, varying by not much more than two percentage points. In this case, the personal exemptions and child credits, along with other features, act as a reasonably approximate adjustment to family size. The AMT's penalties for large families, however, cause tax rates to rise for large families, especially for heads of household (although there are likely only a small fraction of families of these types).

At the $100,000 reference income, the AMT continues to play a role in increasing the penalty on large families. Rates range from 11.23 percent for a childless couple to 19.88 percent for a head of household with six children, a range of eight percent of income. For families without children or families with no more than two children, tax rates are relatively even within head types, although single-headed families pay slightly higher rates than married couples, with or without children, as was the case in the $50,000 and $75,000 reference income levels.

We also consider two very high reference incomes--$250,000 and $500,000. At these levels, singles without children are favored (but only slightly) compared to married couples as the marriage penalty adjustments have shifted in favor of singles. Larger families still tend to pay higher tax rates and, at the $250,000 level, this effect is increased by the AMT. At the highest income levels, the flattening of the rates leads to little differences across family types.

This base case analysis reveals several important features of the tax code that suggest inequitable treatment across family types. First, these discrepancies are extremely pronounced at low income levels, where differentials are around a third of income. That is, the tax system is causing after-tax incomes of the least favored groups to fall by a third relative to that of the most favored groups at the lowest levels. In addition, these discrepancies occur between very common family types ranging from zero to two children.

Second, the types of families favored shift over the income range. At the lowest levels, reflecting the EIC, families with two children are favored; subsequently the largest families are favored, and eventually, the largest families are penalized.

ANALYSIS WITH ADDITIONAL FEATURES OF THE TAX CODE

In this section we consider three additional issues that relate directly to taxation and family size: the dependent care credit, the lack of taxation of imputed income, and examining singles with roommates.

The dependent care provision allows a credit (that declines somewhat with income) against expenses with a maximum of $3,000 for one child and $6,000 for two or more children. Although the dependent care provision is designed to provide a larger credit rate for lower-income families, the provision actually does not benefit the lowest income levels because there is no tax liability or because the credit reduces child credits. In the middle-income classes, as shown in Table 2 and Figure 2, the credit magnifies the dispersion in effective tax rates by providing benefits to families with children, but at the higher income levels is relatively unimportant because of the dollar caps. Thus, the dependent care credit does not play a significant role except in the middle-income classes in causing greater variations in the effective tax rates.

[FIGURE 2 OMITTED]

Table 3 calculates the benefits of excluding child care services in the home by a nonworking spouse (a provision that affects only joint returns). To facilitate comparison with the dependent care credit, we limit the imputation value to the $3,000 and $6,000 caps for the dependent care credit. Note that a case can be made for a larger value of imputed income benefit, since a spouse's home production includes other benefits, some of which may involve monetary savings and some of which permit additional leisure by family members. In keeping with our ability to pay framework, however, we limit our consideration to the case where monetary savings appear most clearly--child care. To implement it, we estimate the benefit of excluding income, while keeping the total of actual and imputed income constant.

As shown in Table 3 and Figure 3, adjusting for this provision does not change the dispersion of tax rates (between the maximum and minimum). At the lowest income levels, it penalizes most families by reducing the earned income credit. At moderate and middle income levels, it tends to benefit joint returns that are more heavily taxed compared to head of household, although it magnifies the differences between couples with and without children.

[FIGURE 3 OMITTED]

The final feature we consider is the possibility of singles who are effectively living together and who potentially enjoy the economies of scale of married couples. These roommate cases could include couples who are effectively in the same situation as married couples but not legally married or more conventional roommates. We treat both as having the same economies as married couples and assume income is split evenly.

[FIGURE 4 OMITTED]

The effect of assuming singles live as roommates reduces effective tax rates at all income levels and leads to increased favoritism at the low and high income levels where the marriage penalty still exists. It equalizes the rates between singles and childless couples in the middle income ranges. Note, however, that we do not assess the effects of a single and a head of household living together.

ANALYSIS WITH MODIFICATION OF ELEMENTS OF THE TAX CODE

In this section, we consider several modifications of the tax code that appear to have important effects on the dispersion of effective tax rates. These modifications include eliminating the AMT, repealing the child credit, and eliminating phaseouts and phase-ins of credits and deductions. We consider each separately and then consider the combination of the last two changes with elimination of the AMT.

As incomes rise, the AMT becomes important. By contrasting Table 5 with Table 1A (bolded tax rates in Table 5 are the rates that would occur without the AMT, but where the AMT is currently paid), and examining Figure 5, which juxtaposes the systems with and without the AMT, we can observe the effect of this provision. At the $50,000 reference income, the AMT has mixed effects on the dispersion of tax rates, and while it narrows the range by raising the tax rates of the favored families with many children who receive significant child credits, it also causes larger head-of-household families to pay higher taxes. At higher income levels (the $75,000 and $100,000 reference incomes), families with many children pay the highest rates because the personal exemption and child credit are not large enough relative to income to adjust for ability to pay. This effect is increased by the AMT, which also limits the evening effect of itemized deductions. At the $75,000 reference income level, for example, tax rates would be relatively even, varying by not much more than two percentage points, but with the AMT the range increases to about six percentage points. In this case, the personal exemptions and child credits, along with other features, act as a reasonably approximate adjustment to family size, but the AMT's penalties for large families create additional dispersions. A similar effect occurs at the $100,000 income. At the very highest income levels, the AMT tends to even out tax rates. Thus, while the AMT contributes to equity across families at the very high income levels, a benefit from its original intention, it creates higher burdens for large families as it reaches down into the less wealthy families.

[FIGURE 5 OMITTED]

Table 6 and Figure 6 show the effects of repealing the $1,000 child credit, a provision initially introduced less than ten years ago (in 1997) and doubled in size in 2001. Although the child credit is not as important at the lowest income levels (because of limits on refundability), it increases the dispersion in tax rates, and plays a role in favoring families with children in the moderate income ranges. For example, in the $15,000 reference income range, tax rates range from zero for a childless couple to -28 percent for a head of household with two children and -22 percent for a couple with two children. Without the credit the rates are -22 percent and -15 percent, respectively. In the $25,000 reference income, rates with the credit range from almost five percent for a single individual to negative eight percent or negative nine percent for certain families with children. Without the credit, most rates for families with children are positive, the lowest rate is negative two percent and there is a fair amount of conformity among married couples regardless of family size. Similarly, at the $50,000 reference income, the rate with the credit ranges from three to nine percent, while the rate without the credit ranges from eight to 12 percent.

[FIGURE 6 OMITTED]

As income levels rise, eventually the credit contributes to more even tax rates. At higher income levels, the flat personal exemption is inadequate to account for family differences and the credit tends to help offset these effects, narrowing the range of tax rates. This effect can be seen in the $75,000 reference income class, where the range of less than nine percent to 15 percent is expanded to a range of slightly over nine percent to 17 percent. Eventually the child credit does not matter because it is phased out.

The child credit, therefore, plays a different role depending on the point in the income distribution. At lower income levels, it acts to increase horizontal inequities as it produces very generous benefits relative to incomes for large families. At higher income levels, it helps to produce more even tax rates, offsetting the inability of the personal exemptions to produce adequate differentials.

The third set of revisions considered is to allow a fully refundable child credit, which affects the lower income levels, and to eliminate the phaseouts at the higher end. These phaseouts include the phasing out of the personal exemption (PEP), the itemized deduction (Pease), and the child credit. The first two are already planned for phaseout and allowing greater refundability has been considered in previous legislation.

As seen in Table 7 and Figure 7, full refundability of the child credit, which affects the lower income brackets, magnifies an already existing favoritism towards families with children. At the lowest reference income level, negative tax rates would exceed 50 percent for most heads of households with two or more children, while tax rates would remain at negative three percent for married childless couples, creating a differential of over half of income. The same effects in direction, although not in magnitude, would lead to tax rates as low as -34 percent in the $15,000 reference income, yielding a differential of a third of income.

[FIGURE 7 OMITTED]

The elimination of phaseouts, by contrast, would narrow the range of effective tax rates at higher incomes. The preservation of itemized deductions, which are, assumed to be proportional to income, helps even out tax rates, as does the preservation, of personal exemptions and child credits.

Table 8 and Figure 8 show the effects of simultaneously repealing the AMT and the child credit. The combination only affects the $50,000, $75,000 and $100,000 categories (as compared to repealing each separately). While repealing both leads to a narrower range for the $50,000 reference income, repealing the child credit does not narrow the range in the other categories (although repealing the AMT does).

[FIGURE 8 OMITTED]

Finally, Table 9 and Figure 9 show the effects of repealing the AMT and eliminating the phaseouts (plus allowing full refundability). As before, the refundability of the child credit greatly magnifies differences at the lower income levels, but at higher levels the repeal of the AMT and the elimination of phaseouts together tend to lead to more even tax rates.

[FIGURES 9-10 OMITTED]

SENSITIVITY ANALYSIS

The tables and graphs in this section compare the extremes of the equivalency index to the base case. Table 10A considers the case where e = 0 and all goods are public. In this case each family would need the same income. This assumption increases the favoritism showed to larger families and extends it up through most of the income scale. Basically, the more public goods there are in the family, the more even are the incomes and the more generous are the benefits for children through the EIC, child credit, head-of-household rate structure, and larger standard deductions for head--of--household returns.

At the other extreme, we treat children as costing the same as adults and assume all goods are private (e = 1). In this case, required income rises proportionally with family size: a family of four needs four times the income of a single person. This assumption reduces, but still maintains, the favoritism towards families at low income levels, but increases the penalties that apply to large families at higher income levels.

CONCLUSIONS

Several findings and policy implications emerge from the analysis. First, the income tax system is characterized at lower income levels with a much more favorable treatment of families with children, especially for families with two or three children. This favorable treatment arises from two features of the tax system: the significantly larger EIC, which increases dramatically with one child and again with two children, and the child credit. The clearest change to emerge from our analysis that would increase horizontal equity, one that even survives the extremes of sensitivity analysis, is a larger EIC for single workers and childless couples.

The analysis at the lower end of the income scale also argues against making the child credit fully refundable if the objective is horizontal equity, another case that could be made even with uncertainty about the equivalency index measure. Expanded refundability has been considered in the past, but it would favor families that are already favorably treated under the income tax.

A revision that is likely to increase horizontal equity at the higher end of the income distribution is the elimination of the phaseouts of itemized deductions, personal exemptions, and child credit. Itemized deductions, because they are relatively proportional to income (at

least across most of the income levels), tend to lead to more even tax rates.

There also appears to be a case for at least preventing the further expansion of the AMT or reducing its scope. The AMT tends to reduce the role played by itemized deductions in leading to more even tax rates, and, for most income classes that it touches, increases the penalties that apply at higher levels to families with children. The AMT does lead to somewhat more even tax rates at the highest income levels, but these are at incomes where very little variation exists. The combination of reducing or eliminating the AMT and eliminating the phaseouts performs well at reducing the variation in tax rates at higher income levels.

The role of the child credit itself is clearly mixed. It contributes to the favoritism towards families with children at the bottom, but reduces the penalties on these larger families at the top. Given the equivalency index approach, the provision is fundamentally flawed because it does not vary with income. If the equivalency index were known with certainty, we could obtain perfect horizontal equity by varying exempt levels and rate brackets with family characteristics. If family A needs 250 percent of the income of family B, A's exempt amounts and rate brackets would be 250 percent of the size of B's. At higher income levels, where itemized deductions are more likely to rise with income, there is a built--in adjustment. But at lower income levels, the increased tax benefits for children rise too quickly, while at higher levels, they rise too slowly.

Acknowledgments

The views expressed in this paper represent those of the authors and not necessarily those of the Congressional Research Service or the Government Accountability Office.

REFERENCES

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

Fields, Jason. "American Families and Living Arrangements: 2003." Current Population Reports. P20-553. Washington, D.C.: U.S. Census Bureau, 2004.

Hoffman, Saul D., and Laurence S. Seidman. Helping Working Families: The Earned Income Tax Credit. Kalamazoo: W. E. Upjohn Institute for Employment Research, 2003.

Sailer, Peter, Ellen Yau, and Victor Rehula. "Income by Gender and Age from Information Returns." IRS Statistics of Income Bulletin (Winter, 2001): 83-102.

Scott, Christine. "The Earned Income Tax Credit (EITC): An Overview." Congressional Research Service Report RL31768. Washington, D.C.: Library of Congress, 2006.

Weber, Michael E. "The 1993 Family Cross Section File: Combining Parents and Dependents into Family Units." IRS Statistics of Income Bulletin (Fall, 1998): 231-44.

Jane Gravelle

Congressional Research

Service, Washington,

D.C. 20540

Jennifer Gravelle

Government

Accountability Office,

Washington, D.C.

20548

TABLE 1A
DISTRIBUTION OF FAMILY TYPES ACROSS INCOME CLASSES, 2003

                              Distribution within Income Classes

Income       Total                 Head of    Married w/   Married No
Class     Distribution   Single   Household    Children     Children

<10          19.95       75.51      12.64        4.11         6.40
10-15         9.20       57.47      23.48        7.44        10.01
15-25        16.20       47.73      25.84       11.93        12.18
25-50        25.26       41.25      17.62       13.56        24.76
50-75        13.32       24.65       8.40       28.38        36.98
75-100        7.32       13.34       6.91       36.88        44.73
100-200       6.81       10.67                  31.99        54.10
200-500       1.53       11.01                  31.47        54.43

* Observations too small to report

Source: Internal Revenue Service Statistics of Income.

TABLE 1B
INCOME LEVELS

                  Reference Income Level

Children   10,000    15,000    25,000    50,000

                     Married Filing Joint

0           10,000    15,000    25,000      50,000
1           12,338    18,507    30,844      61,688
2           14,498    21,747    36,246      72,491
3           16,529    24,792    41,321      82,641
4           18,456    27,685    46,141      92,282
5           20,302    30,453    50,754     101,509
6           22,078    33,117    55,194     110,383

                   Single/Head of Household

0            6,156     9,233    15,389      30,779
1            8,925    13,387    22,312      44,623
2           11,361    17,042    28,403      56,806
3           13,590    20,386    33,976      67,952
4           15,672    23,508    39,180      78,361
5           17,641    26,462    44,103      88,206
6           19,520    29,280    48,800      97,600

                   Reference Income Level

Children   75,000    100,000   250,000   500,000

                     Married Filing Joint

0           75,000   100,000   250,000     500,000
1           92,533   123,377   308,422     616,884
2          108,737   144,982   362,455     724,910
3          123,962   165,283   413,207     826,414
4          138,423   184,564   461,411     922,822
5          152,263   203,017   507,543   1,015,086
6          165,583   220,777   551,942   1,103,883

                   Single/Head of Household

0           46,168    61,557   153,893     307,786
1           66,935    89,247   223,117     446,235
2           85,210   113,613   284,032     568,063
3          101,928   135,904   339,760     679,520
4          117,451   156,721   391,803     783,606
5          132,309   176,412   441,030     882,059
6          146,401   195,201   488,002     976,003

TABLE 1C
EFFECTIVE TAX RATES FOR BASE CASE (BC)

                   Reference Income
          (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000

M 0        -2.87         0      3.44      8.32
M 1        -23.2    -17.94     -0.73      6.56
M 2       -33.97    -22.44      -2.4      5.29
M 3       -31.43    -18.94     -3.47      4.28
M 4       -27.52    -16.33     -4.79      3.45
M 5       -24.47    -14.29     -0.21      2.95
M 6       -22.01    -12.65     -7.45      3.77
S 0        -6.48     -0.96      4.67      8.95
HH 1      -29.83    -22.56     -6.86      6.14
HH 2      -39.21    -27.83     -7.89      4.72
HH 3      -35.24    -22.27     -5.12      3.72
HH 4      -30.8     -18.51     -5.97      4.87
HH 5      -26.68    -15.77     -7.35      6.34
HH 6      -23.53    -13.67     -8.55      7.39

                   Reference Income
          (Married Filing Joint No Children)

Family   75,000    100,000   250,000   500,000

M 0         9.45     11.23     19.41     24.75
M 1         8.62     11.91     21.81     24.93
M 2         8.58     12.52     23.35     25.04
M 3         9.04     14.01     24.43     25.12
M 4        10.30     15.46     24.67     25.37
M 5        11.37     16.74     24.76     25.64
M 6        12.22     17.73     24.84     25.86
S 0        10.54     12.78     17.78     24.16
HH 1        8.89     12.38     21.28     24.63
HH 2        9.7      13.6      23.6      24.86
HH 3       11.19     15.69     24.30     25
HH 4       12.60     17.64     24.49     25.29
HH 5       13.82     18.94     24.62     25.63
HH 6       15.07     19.88     24.72     25.89

TABLE 2
EFFECTIVE TAX RATES WITH MAXIMUM DEPENDENT CARE CREDIT (DCC)

            Reference Income (Married Filing Joint No Children)

Family    25,000    50,000    75,000   100,000   250,000   500,000

M 0         3.44      8.32      9.45     11.23     19.41     24.75
M 1        -3.45      5.59      7.97     11.42     21.62     24.83
M 2        -6.11      3.63      7.48     11.69     23.02     24.87
M 3        -6.95      2.83      8.07     13.29     24.14     24.97
M 4        -7.65      2.15      9.43     14.81     24.41     25.24
M 5        -8.57      1.77     10.58     16.15     24.52     25.53
M 6        -9.62      2.69     11.50     17.18     24.62     25.75
S 0         4.67      8.95     10.54     12.78     17.78     24.16
HH 1      -10.72      4.79      7.99     14.07     23.96     24.5
HH 2      -12.13      2.61      8.29     12.54     21.72     24.65
HH 3       -9.53      1.98      9.68     14.81     23.95     24.82
HH 4       -9.34      2.67     11.39     16.87     24.18     25.14
HH 5      -10.07      4.47     12.81     18.26     24.35     25.5
HH 6      -11.01      5.77     14.25     19.37     24.48     25.77

TABLE 3
EFFECTIVE TAX RATES WITH IMPUTED INCOME (IMINC)

      Reference Income (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000

M 0        -2.87      0         3.44      8.32
M 1       -21.58    -18.04     -3.26      5.99
M 2       -23.45    -23.51     -7.54      4.32
M 3       -25.48     -20.4     -5.99      3.43
M 4       -25.02    -17.64      -6.3      2.69
M 5       -24.11    -15.49     -7.42      2.05
M 6       -23.38    -13.75     -8.56      2.43
S 0        -6.48     -0.96      4.67      8.95
HH 1      -29.83    -22.56     -6.86      6.14
HH 2      -39.21    -27.83     -7.89      4.72
HH 3      -35.24    -22.27     -5.12      3.72
HH 4       -30.8    -18.51     -5.97      4.87
HH 5      -26.68    -15.77     -7.35      6.34
HH 6      -23.53    -13.67     -8.55      7.39

      Reference Income (Married Filing Joint No Children)

Family    75,000   100,000   250,000   500,000

M 0         9.45     11.23     19.41     24.75
M 1         7.99     11.31     21.5       24.8
M 2         7.51     11.51     22.82     24.83
M 3         7.85     12.88     23.96     24.93
M 4         9.05     14.33     23.84     25.19
M 5         10.2     15.71     24.46     25.48
M 6        11.08     16.78     24.56     25.71
S 0        10.54     12.78     17.78     24.16
HH 1        8.89     12.38     21.28     24.63
HH 2        9.7      13.6      23.6      24.86
HH 3       11.19     15.69     24.3      25
HH 4        12.6     17.64     24.49     25.29
HH 5       13.82     18.94     24.62     25.63
HH 6       15.07     19.88     24.72     25.89

TABLE 4
EFFECTIVE TAX RATES WITH SINGLES LIVING AS ROOM MATES (RM)

      Reference Income (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000

M 0        -2.87      0         3.44      8.32
M 1       -23.2     -17.94     -0.73      6.56
M 2       -33.97    -22.44     -2.4       5.29
M 3       -31.43    -18.94     -3.47      4.28
M 4       -27.52    -16.33     -4.79      3.45
M 5       -24.47    -14.29     -6.21      2.95
M 6       -22.01    -12.65     -7.45      3.77
S 0        -7.65     -4.33      3.44      8.32
HH 1      -29.83    -22.56     -0.86      6.14
HH 2      -39.21    -27.83     -7.89      4.72
HH 3      -35.24    -22.27     -5.12      3.72
HH 4      -30.8     -18.51     -5.97      4.87
HH 5      -26.68    -15.77     -7.35      6.34
HH 6      -23.53    -13.67     -8.55      7.39

      Reference Income (Married Filing Joint No Children)

Family    75,000   100,000   250,000   500,000

M 0         9.45     11.23     19.41     24.75
M 1         8.62     11.91     21.81     24.93
M 2         8.58     12.52     23.35     25.04
M 3         9.04     14.01     24.43     25.12
M 4        10.3      15.46     24.67     25.37
M 5        11.37     16.74     24.76     25.64
M 6        12.22     17.73     24.84     25.86
S 0         9.45     11.23     16.73     22.44
HH 1        8.89     12.38     21.28     24.63
HH 2        9.7      13.6      23.6      24.86
HH 3       11.19     15.69     24.3      25
HH 4       12.6      17.64     24.49     25.29
HH 5       13.82     18.94     24.62     25.63
HH 6       15.07     19.88     24.72     25.89

TABLE 5
EFFECTIVE TAX RATES WITH REPEALING THE AMT (NAMT)

      Reference Income (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000

M 0        -2.87      0         3.44      8.32
M 1       -23.2     -17.94     -0.73      6.56
M 2       -33.97    -22.44     -2.4       5.29
M 3       -31.43    -18.94     -3.47      4.28
M 4       -27.52    -16.33     -4.79      3.45
M 5       -24.47    -14.29     -6.21      2.75
M 6       -22.01    -12.65     -7.45      2.27
S 0        -6.48     -0.96      4.67      8.95
HH 1      -29.83    -22.56     -6.86      6.14
HH 2      -39.21    -27.83     -7.89      4.72
HH 3      -35.24    -22.27     -5.12      3.72
HH 4      -30.8     -18.51     -5.97      3.8
HH 5      -26.68    -15.77     -7.35      4.08
HH 6      -23.53    -13.67     -8.55      4.18

      Reference Income (Married Filing Joint No Children)

Family    75,000   100,000   250,000   500,000

M 0         9.45     11.23     17.16     22.86
M 1         8.62     11.91     19.68     23.9
M 2         8.58     12.52     20.98     24.56
M 3         9.04     13.01     21.7      25.03
M 4         9.37     13.4      22.4      25.37
M 5         9.6      13.78     22.94     25.64
M 6         9.76     14.09     23.37     25.86
S 0        10.54     12.78     17.78     22.04
HH 1        8.89     12.38     18.33     22.98
HH 2        9.7      13.38     20.42     24.13
HH 3       10.32     13.83     21.62     24.82
HH 4       10.69     14.2      22.3      25.29
HH 5       10.88     14.64     22.91     25.63
HH 6       10.96     15.34     23.44     25.89

TABLE 6
EFFECTIVE TAX RATES WITH REPEALING CHILD CREDIT (NCC)

               Reference Income (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000    75,000   100,000

M 0        -2.87      0         3.44      8.32      9.45     11.23
M 1       -21.58    -12.54      2.51      8.18      9.7      12.15
M 2       -30.35    -15.03      3.12      8.04     10.42     12.69
M 3       -26.42    -10.59      3.8       7.91     10.89     14.14
M 4       -21.46     -7.29      3.88      7.79     12.14     15.6
M 5       -17.59     -4.71      3.64      7.88     13.24     16.89
M 6       -14.48     -2.64      3.42      9.16     14.15     17.93
S 0        -6.48     -0.96      4.67      8.95     10.54     12.78
HH 1      -23.2     -19.88     -2.38      8.38     10.38     12.66
HH 2      -38.73    -22.43     -0.85      8.25     11.4      13.64
HH 3      -32.38    -15.23      3.71      8.16     12.47     15.69
HH 4      -26.33    -10.44      4.24      9.05     13.98     17.64
HH 5      -21.04     -7         3.99     10.7      15.31     18.94
HH 6      -16.98     -4.3       3.74     11.7      16.71     19.88

TABLE 7
EFFECTIVE TAX RATES WITH NO PHASEOUTS AND FULL CHILD CREDIT (NPP FCC)

      Reference Income (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000

M 0        -2.87      0         3.44      8.32
M 1       -29.68    -17.94     -0.73      6.56
M 2       -44.14    -24.22     -2.4       5.29
M 3       -44.57    -22.69     -3.47      4.28
M 4       -43.13    -21.73     -4.79      3.45
M 5        12.22    -21.13     -0.21      2.95
M 6        11.66    -20.75     -7.45      3.73
S 0        -6.48     -0.96      4.67      8.95
HH 1      -41.03    -27.35     -0.86      6.14
HH 2      -56.33    -34.17     -7.89      4.72
HH 3      -54.45    -29.94     -5.12      3.72
HH 4      -51.15    -27.46     -5.97      3.94
HH 5      -49.38    -25.9      -7.35      5.03
HH 6      -47.72    -24.79     -8.55      5.82

      Reference Income (Married Filing Joint No Children)

Family    75,000   100,000   250,000   500,000

M 0         9.45     11.23     18.97     24.15
M 1         8.62     11.34     20.94     24.12
M 2         8.58     11.31     22.17     24.09
M 3         8.47     12.14     23.02     24.07
M 4         9.25     13.23     23.23     24.04
M 5         9.89     14.15     23.18     24.01
M 6        10.35     14.88     23.13     23.99
S 0        10.54     12.78     17.69     23.71
HH 1        8.89     11.54     20.47     23.84
HH 2        9.02     11.88     22.39     23.88
HH 3        9.53     13.49     22.94     23.9
HH 4       10.58     14.95     22.94     23.89
HH 5       11.53     15.93     22.92     23.89
HH 6        12.6     16.66     22.91     23.88

TABLE 8
EFFECTIVE TAX RATES WITH REPEALING BOTH CHILD CREDIT AND AMT (NCC NAMT)

      Reference Income (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000

M 0        -2.87      0         3.44      8.32
M 1       -21.58    -12.54      2.51      8.18
M 2       -30.35    -15.03      3.12      8.04
M 3       -26.42    -10.59      3.8       7.91
M 4       -21.46     -7.29      3.88      7.79
M 5       -17.59     11.71      3.64      7.67
M 6       -14.48     -2.64      3.42      7.66
S 0        -0.48     -0.96      4.67      8.95
HH 1      -23.2     -19.88     -2.38      8.38
HH 2      -38.73    -22.43     -0.85      8.25
HH 3      -32.38    -15.23      3.71      8.16
HH 4      -26.33    -10.44      4.24      8.65
HH 5      -21.04     -7         3.99      8.95
HH 6      -16.98     -4.3       3.74      9.15

      Reference Income (Married Filing Joint No Children)

Family    75,000   100,000   250,000   500,000

M 0         9.45     11.23     17.16     22.86
M 1         9.7      12.15     19.68     23.9
M 2        10.42     12.69     20.98     24.56
M 3        10.89     13.13     21.7      25.03
M 4        11.21     13.54     22.4      25.37
M 5        11.47     13.93     22.94     25.64
M 6        11.7      14.29     23.37     25.86
S 0        10.54     12.78     17.78     22.04
HH 1       10.38     12.66     18.33     22.98
HH 2       11.4      13.42     20.42     24.13
HH 3       11.94     13.83     21.62     24.82
HH 4       12.27     14.2      22.3      25.29
HH 5       12.47     14.64     22.91     25.63
HH 6       12.6      15.34     23.44     25.89

TABLE 9
EFFECTIVE TAX RATES WITH NO PHASEOUTS, FULL
CHILD CREDIT, AND NO AMT (NPP FCC AMT)

      Reference Income (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000

M 0        -2.87      0         3.44      8.32
M 1       -29.68    -17.94     -0.73      6.56
M 2       -44.14    -24.22     -2.4       5.29
M 3       -44.57    -22.69     -3.47      4.28
M 4       113.13    -21.73     -4.79      3.45
M 5       -42.22    -21.13     -6.21      2.75
M 6       111.66    -20.75     -7.45      2.22
S 0        -0.48     -0.96      4.67      8.95
HH 1      -41.03    -27.35     -0.86      6.14
HH 2      -56.33    -34.17     -7.89      4.72
HH 3      -54.45    -29.94     -5.12      3.72
HH 4      -51.15    -27.46     -5.97      3.54
HH 5      -49.38    -25.9      -7.35      3.29
HH 6      117.72    -24.79     -8.55      3

      Reference Income (Married Filing Joint No Children)

Family    75,000   100,000   250,000   500,000

M 0         9.45     11.23     17.13     21.66
M 1         8.62     11.34     18.1      23.29
M 2         8.58     11.31     18.67     22.83
M 3         8.47     11.23     19.04     23.12
M 4         8.32      11.2     19.36     23.33
M 5         8.16     11.23     19.66     23.48
M 6         7.98     11.22     19.89     23.6
S 0        10.54     12.78     17.69     21.17
HH 1        8.89     11.54     17.21     21.55
HH 2        9.02     11.66     18.32     22.41
HH 3        9        11.62     18.93     22.9
HH 4        8.86     11.59     19.31     23.21
HH 5        8.69     11.66     19.56     23.43
HH 6        8.5      11.67     19.87     23.58

TABLE 10A
EFFECTIVE TAX RATES ASSUMING e = 0 (ALL GOODS PUBLIC)

      Reference Income (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000

M 0        -2.87      0         3.44      8.32
M 1       -26.62    -21.75     -6.97      5.36
M 2       -40       -33.33    -17.45      2.4
M 3       -40       -33.33    -18.73     -0.56
M 4       -40       -33.33    -18.73     -3.52
M 5       -40       -33.33    -18.73     -0.48
M 6       -40       -33.33    -18.73     -9.32
S 0         0.46      4.53      8.32     11.23
HH 1      -26.62    -21.08     -3.16      6.74
HH 2      -40       -32.45    -13.41      3.78
HH 3      -40       -32.45    -17.05      0.82
HH 4      -40       -32.45    -17.05     -2.15
HH 5      -40       -32.45    -17.05     -5.11
HH 6      -40       -32.45    -17.05     -8.07

      Reference Income (Married Filing Joint No Children)

Family    75,000   100,000   250,000   500,000

M 0         9.45     11.23     19.41     24.75
M 1         7.47      9.43     19.41     24.75
M 2         5.5       7.63     19.41     24.75
M 3         3.53      5.83     19.41     24.75
M 4         1.55      4.09     19.41     24.75
M 5        -0.42      2.64     19.41     24.75
M 6        -2.39      1.64     19.41     24.75
S 0        13.99     15.45     22.44     24.75
HH 1       10.03     13.4      22.44     24.75
HH 2        7.63     11.85     22.44     24.75
HH 3        5.23      10.5     22.44     24.75
HH 4        3.05       9.5     22.44     24.75
HH 5        1.72       8.5     22.44     24.75
HH 6        0.39       7.5     22.44     24.75

TABLE 10B
EFFECTIVE TAX RATES ASSUMING e = 1, p = 1
(NO PUBLIC GOODS, CHILDREN EQUIVALENT TO ADULTS)

      Reference Income (Married Filing Joint No Children)

Family    10,000    15,000    25,000    50,000

M 0        -2.87      0         3.44      8.32
M 1       -21.75    -10.63      2.55      7.47
M 2       -24.93     -9.37      2.4       7.63
M 3       -18.73     -4.93      1.89      9.16
M 4       -14.6      -5.24      1.55     11.73
M 5       -11.65     -5.61      1.31     13.82
M 6       -10.88     -5.92      1.64     15.94
S 0        -7.65     -4.33      3.44      8.32
HH 1      -26.62    -21.08     -3.16      6.74
HH 2      -32.45    -18.35      1.51      7.63
HH 3      -22.82    -10.39      0.82      10.5
HH 4      -17.05     -6.76      0.62     13.37
HH 5      -13.19     -6.97      1.72     16.21
HH 6      -10.44     -7.12      4.48     18.28

      Reference Income (Married Filing Joint No Children)

Family    75,000   100,000   250,000   500,000

M 0         9.45     11.23     19.41     24.75
M 1        10.68     13.47     23.64     25.06
M 2        13.07     16.69     24.75      25.6
M 3        15.82     19.41     24.94     26.15
M 4        18.15     21.52     25.06     26.52
M 5        20.01     23.04     25.21     26.78
M 6        21.52     24.17     25.6      26.98
S 0         9.45     11.23     16.73     24.75
HH 1       10.03     13.4      22.44     25.06
HH 2       13.45     17.05     24.43     25.6
HH 3       17.05     20.22     24.75     26.15
HH 4       19.58     22.44     24.94     26.52
HH 5       21.37     24.06     25.15     26.78
HH 6       22.9      24.34     25.61     26.98