The negative effect of the marriage penalty tax on American society.

Feucht, Frederick J. ; Smith, L. Murphy ; Strawser, Robert H. 等

INTRODUCTION

The marriage penalty tax (MPT) has been a topic of considerable discussion and debate by politicians, academicians, researchers, legislators, and, of course, individual taxpayers ever since Congress enacted the 'married filing jointly' status in 1913. MPT occurs when the total tax liability incurred by a married couple on their combined income exceeds the amount owed had they filed as separate/single individuals. To engender the public trust, tax policy must be both fair and equitable. Policy makers must not only consider equity but also political and social implications as well in the formulation and subsequent modification of tax law.

The objective of this study is to investigate empirically the impact of changes in tax law to determine whether, in fact, legislation truly accomplishes the intentions of Congress and the President. An important caveat in this regard is that Presidential and Congressional intentions, like tax laws, are subject to continual change as a result of changes in the political party that controls the White House and Congress. In 2006, control of Congress switched from Republican to Democratic control (Cook 2006).

To fully evaluate the impact of the new tax law, its distributional effects on each income class of taxpayers must be accurately identified. The results of this study will assist policy makers and tax researchers in assessing the net impact of changes in tax policy on income redistribution regarding MPT in the U.S. We extend previous MPT literature by empirically assessing the impact of changes in tax law on MPT in an attempt to determine whether the higher tax burden on married couples is reduced or eliminated. Further, the effects on and consequences of MPT on society as a whole are considered.

On June 7, 2001, President Bush signed into law The Economic Growth and Tax Relief Reconciliation Act of 2001 (EGTRRA 2001) asserting that, among other changes, this Act was intended to reduce MPT. Relief was administered in two forms--the phase-in of an increase in the standard deduction for married couples and expansion of the income subject to the 15 percent rate to an amount equal to twice that of single taxpayers. Subsequently, the phase-in schedule for the increase in the standard deduction was delayed to 2005 (at 174% of the single standard deduction) and gradually increasing to 190% in 2008. The expanded 15% rate bracket for married couples filing joint returns increased from 180% of the

single bracket in 2005 and to 200% in 2008.

On May 28, 2003, in an effort to spur a lagging economy, President Bush signed the Jobs and Growth Tax Relief Reconciliation Act of 2003 (JGTRRA 2003) into law. JGTRRA 2003 accelerated many of the provisions of EGTRRA 2001, increased the childcare tax credit, temporarily doubled the standard deduction for married couples, and attempted to address the inequity of MPT. In an address to the Nation, President George W. Bush stated: "My tax cut plan is not just about productivity, it is about people. Economics is more than narrow interests or organized envy. A tax plan must apply market principles to the public interest. And my plan sets out to make life better for average men, women, and children. The current tax code is full of inequities. Many single moms face higher marginal tax rates than the wealthy. Couples frequently face a higher tax burden after they marry. High marginal tax rates act as a tollgate, limiting the access of low and moderate-income earners to the middle class. The current tax code frequently taxes couples more after they get married. This marriage tax contradicts our values and any reasonable sense of fairness" [emphasis added] (White House 2003).

According to Republican Representative Nancy Johnson of Connecticut, the MPT is especially onerous to women. Representative Johnson said: "I'll tell you our tax code and the penalty that it levies on married couples is not only cruel and unusual but it strikes at the dignity of women" (PBS 2000). The MPT has historically been a divisive partisan issue in which Republicans favor total elimination of the MPT, while Democrats are mostly opposed (Stevenson 2000, CNN 2000, PBS 2000). Occasionally, some Democrats align with Republicans to support tax reductions, including elimination of the MPT (Allen 2008). Not surprisingly, demographic data based on the 2004 election shows that shows that married people, especially parents with children at home, strongly support Republican positions. Exit polling data indicated that President Bush led by 57%42% among all married Americans, and 59%-40% among married people with children (Pew Research Center 2005).

With the enactment of the Jobs and Growth Tax Relief Reconciliation Act of 2003, President Bush indicated that the burden of MPT in the American tax system would be removed or at least largely eliminated. As Kinney (2003) appropriately stated, "In policy related research, if the government says something is true, we, as accounting researchers, can test to see if it actually is true."

Research on tax policy has examined the impact of such issues as changes in capital gains taxes or depreciation rules and their impact on businesses and individuals. It is widely recognized that public policy implemented via tax law affects decision-making by business managers and individuals and thereby affects the economy. Perhaps not so obvious is the effect on issues such as marriage. Social researchers have found that marriage decisions have significant consequences on individuals, their children (if they are or become parents), and our society at large. The impact of marriage on children is highly researched (e.g. Demuth and Brown 2004, Deleire and Kalil 2002, Hoffman and Johnson 1998). Research also examines the impact of marriage on marriage partners, whether parents or not (Colson and Fickett 2005, Massachusetts 2004, Morse 2001, Popenoe and Whitehead 2000). A stream of research also examined the impact of tax law on marriage (e.g. Whittington and Alm 2001; Cook et al. 2001; Alm and Whittington 1995, 1996, 1999; Moffit et al. 1998; Gelardi 1996; Feenberg and Rosen 1995).

This study utilizes aggregate data from the Statistics of Income (SOI) department of the Internal Revenue Service (IRS 2004). This data consists of actual income tax returns filed by married filing joint taxpayers for the year 2000. Additionally, census data maintained by the Census Bureau's year 2000 Current Population Survey (CPS) is employed to allocate consistently the various items of income, expenses, dependency deductions, credits, and other items (Census Bureau 2000), allowing us to calculate and compare MPT before and after changes in tax law. Research questions addressed in this study are designed to examine whether tax legislation does, in fact, remove or largely eliminate MPT.

This study is of significance to legislators, policy makers, researchers and, of course, to individual taxpayers. Additional exploration and discussion of these changes is strongly encouraged in order to derive a comprehensive view of their impact on taxpayers' marital status. The availability of current income tax data, significant shifts in spousal employment trends, and recent changes in tax legislation suggest that knowledge of MPT must be updated.

HISTORY OF MARRIAGE PENALTY TAX

MPTs (or marriage bonus tax--MBT, i.e., lower tax) arise when the total tax liability incurred by a married couple on their combined income is greater (less) than that owed had they filed as single/separate individuals. The question of some policy makers, especially the U.S. Congress, is that if MPT were eliminated (thus decreasing federal tax revenues), should not the marriage bonus tax also be eliminated? Such a perspective may be misguided, if indeed the concern is truly about overall federal revenues and outlays. The biggest so-called marriage "bonus" (reduced) tax goes to a married couple in which only one spouse is working, which may mean that one spouse gave up a job to be a full-time caregiver for children. If the couple were to nullify their marriage, the MBT (decrease in federal revenues) would, in most cases, likely be more than offset by government transfer programs (increase in federal outlays such as EITC, welfare, food stamps, etc.) that would be due to the non-working person (former spouse). Thus, eliminating the MPT should not automatically call for eliminating the so-called "bonus," at least not if the purpose is to improve federal cash flows. In any case, for the following analyses, the net MPT (or bonus) will be evaluated. Precisely determining net MPT (or MBT) is a complex problem. Researchers examining MPT/MBT questions have found that several assumptions must be made regarding the division of income to estimate the tax liability a married couple would have incurred had they been permitted to file as two single/separate individual taxpayers. Numerous factors contribute to the disparity in tax liabilities between individuals filing as single versus married filing jointly. Tax rates, tax brackets, allowable standard deductions, childcare credits, and other factors all cause a shift in tax liabilities. The intricacies of the earned income tax credit and the various phase-in/out thresholds further complicate the analysis. Essentially, while holding income and investment decisions constant, research has indicates that marriage itself accounts for a considerable disparity in tax liability. While married couples may encounter differences in overall cash-flow outlays or spending compared to singles, this study will limit its investigation to the disparity (i.e., MPT), which arises solely because of marital status. In addition, the social costs associated with the MPT will be considered.

McIntyre and McIntyre (1999) confirm that MPTs have been in place since the adoption of the federal income tax in 1913. At that time, a personal exemption of $3,000 was allowed to a single person and $4,000 (or $2,000 each) to a married couple. Thus, a single individual received a personal tax exemption which was $1,000 or fifty-percent greater than that of a married individual.

In 1948, most significant MPTs were temporarily eliminated with the adoption of full income splitting for married couples. The Revenue Act of 1969 reintroduced MPTs by abandoning full income splitting in favor of marital joint filing with partial income splitting. The Tax Reform Act of 1986 significantly reduced most MPTs, primarily by a substantial reduction in tax rates. The Revenue Reconciliation Acts of 1990 and 1993 increased MPTs for high-income married couples while lowering MPTs for certain low-income married couples with children. However, the 1990 and 1993 acts increased MPTs for many low-income couples with children through the imposition of limitations of income on earned income tax credits (EITC).

TAX RATE SCHEDULES AND STANDARD DEDUCTION

The U.S. employs a voluntary income tax system based on progressive rates; tax rates increase as taxable income increases. U.S. taxes subject four classes of taxpayers (married filing jointly, married filing separately, single, and head of household) to progressive gradient levels, or brackets, of tax rates.

Additionally, in determining taxable income, taxpayers are allowed a deduction for specific expenses (itemized deductions) for medical costs, certain taxes, mortgage interest, charitable contributions, casualty losses, job expenses, and certain expenses incurred in connection with the production or collection of income. For taxpayers who do not have sufficient expenses, a standard deduction, a different amount to each class, is allowed to each group of taxpayer in lieu of itemized deductions. Thus, taxpayers may take a deduction for their actual itemized expenses or for the standard deduction, whichever is greater. The standard deduction allowed to the taxpayers is regarded as an efficiency technique useful in minimizing tax administration since this deduction is deemed to be de minimis in nature. Exhibit 1 shows standard deductions for tax years pre- and post-JGTRRA.

In general, MPT arises because of disparity in the bracketed amounts of the graduated tax rate schedules as well as in the standard deductions allowed for married vs. non-married taxpayers, ceteris paribus. Since tax rate brackets and standard deductions for married couples filing jointly are not exactly twice those allowed for single individuals, a difference that tends to create MPT for the couple filing as married jointly occurs. JGTRRA 2003 attempted to mitigate this disparity, at least temporarily, in the lower brackets by setting the brackets of the married joint tax rates equal to twice that of single taxpayers for the initial two tax rate levels (10% and 15%). However, disparity is once again evident when the subsequent four tax rate brackets for each class of taxpayer are compared.

For 2003 and 2004, married filing joint taxpayers were allowed a deduction equal to twice that of the single person. In 2005, the deduction decreased to 174% but increases to 200% in year 2009. These reductions in MPT will 'sunset' or expire after the year 2010. This sunset provision may be intended to permit the evaluation of the impact of the tax law prior to proposing a permanent substantial reduction in or eventual elimination of MPT. This research may be useful in resolving this issue.

PRIOR EMPIRICAL RESEARCH

Historically, interpretations of MPT (MBT) vary greatly. McIntyre and McIntyre (1999) state that total elimination of MPT would reduce federal tax revenues by approximately $42 billion per year (stated in 1998 dollars). Bull et al. (1999) find that MPTs (MBT) were more easily defined than accurately measured. Brozovsky and Cataldo (1994) state that there is a trend toward increasing marginal tax rates and, should it continue, will eventually lead to subsequent increases in MPT. Wiggins et al. (1986) determine that MPTs are incurred primarily by middle- and upper-income couples. Yet recent research shows that MPTs are incurred by all classes of taxpayers and may have the greatest impact on low-income couples (Cook et al. 2001; Smith et al. 1999).

Research on MPT has expanded to several related areas; the impact of tax and transfer programs on social behavior, the effects of changes in tax policy on individual behavior, and attempts to quantify the true costs to society. Using time-series analysis, Alm and Whittington (1995) explored changes in MPT affecting the probability of marriage or divorce. Smith et al. (1999) examined the effect of U.S. tax policies on marital stability--more specifically, how MPT affects American society. Gelardi (1996) investigated the timing of marriages in relation to changes in tax law and found that proposed changes prompted a significant decrease in the number of marriages occurring in the final half of one year and an unusually high increase in the initial half of the subsequent year, concluding that changes in tax laws do affect individual behavior.

Since marital filing status for federal income tax purposes is determined on the final day of each calendar year, couples planning to marry may plan for the event to take place pre- or post-January 1. Strategically, if only one of the two earns income, the marriage would benefit by occurring on or before December 31 whereas, if both earn income, MPT could be avoided by a post-January 1 wedding, ceteris paribus.

Researchers agree that numerous factors affect the imposition of MPT (MBT). Among these factors are current tax rate schedules, standard deduction, child care credit, earned income tax credit, alternative minimum tax, and individual family situations.

To quantify the net effect of MPT throughout the various changes in tax law, prior studies have empirically calculated average MPTs and benefits (a benefit of lesser tax paid by married couples vs. that paid by two single taxpayers, ceteris paribus) among all income groups per bracketed amount. After comparing the net MPT with the net MBT, a "real" net MPT was calculated. Exhibit 2 reports a summary of the net MPT for the period 1980-1999.

Consistent with prior research, Holtzblatt and Reberlein (2000) measured the effect of the earned income tax credit (EITC) on MPT (MBT) by employing a 'resource pooling' measurement method using the Treasury Department's Individual Tax Model (a micro-simulation model based on the Statistics of Income sample of tax returns for 1995). The effect on the married filing jointly income tax returns for the 2000 tax year (stratifying the results into five classes by adjusted gross income) was determined and, subject to various assumptions, it was estimated that MPTs would increase by at least $3.1 billion (10.4 percent) while MBTs would decrease by $439 million (1.5 percent). The most notable effect on MPTs occurs in situations where couples are ineligible for EITC because their combined incomes exceed $30,000 (maximum phase-out).

Feenberg and Rosen (1995) utilized a Tax Simulation Model (TAXSIM) maintained by the National Bureau of Economic Research to test a stratified random sample (96,589) of tax returns filed in 1989 and 'aged' the data using a nominal per capita income growth, consistent with Congressional Budget Office (CBO) predictions, to 1993/1994 levels. Further, since the original data did not indicate the division of the earnings by couples, family earnings were divided between the husband and wife using assumptions and data obtained from the March 1990 Current Population Survey (CPS) thus separating the joint liability of the couple into two separate liabilities. Assumptions also were made for the division of passive income, dependents, and other factors.

Using 1996 tax data, the U.S. CBO (1997) found that among married couples, a significant increase in two wage-earner households occurred in the last two decades. CBO also found that there was increasing equality of the incomes of marriage partners, particularly in households with higher incomes, contributing to a substantial increase in the share and magnitude of couples incurring MPTs. CBO estimated that 42 percent of couples had an average MPT of roughly 2 percent of adjusted gross income (range 7.6%-1.6%) with higher estimates calculated if itemized deductions and dependents were assigned to the higher income spouse (based on a hypothetical divorce).

Prior research also considers the effect of changes in tax policy on the behavior of individuals and their decision to marry, such as, how MPT affects single couples contemplating marriage (or possible postponement--temporarily or permanently) and/or married couples contemplating divorce. Moffit et al. (1998) report strong evidence that economic considerations in the welfare system favor unmarried individuals encouraging low-income partners to cohabitate rather than marry, resulting in possible socially detrimental consequences on children, as discussed later in this paper.

Sjoquist and Walker (1995) examined the impact of MPT on the rate and timing of marriage. Alm and Whittington (1995, 1996, and 1999) concluded that an increase in MPT of $465 decreases the probability of an initial marriage for low-income women by approximately 3% whereas, for wealthy professional couples, an increase in MPT (up to 99%) could decrease the probability of initial marriages by as much as 23%.

Researchers also attempt to identify and quantify the "true" long-term cost to society associated with changes in public policy and measure the impact of these changes on human behavior. Smith et al. (1999) provide an analysis of the impact of societal ethics, cultural trends, divorce law and public policies on marriage. They posit that public policy (i.e., tax and transfer programs) may affect marital stability and state an expectation that government policy would be more pro-child, pro-marriage (or at least neutral to marriage) and advance legislation that promotes marriage. It is estimated that MPTs cost American married couples approximately $33 billion in the year 2000, an explicit tax cost to married taxpayers and their families (Olshewsky et al. 2001).

Researchers have attempted to ascertain the 'true' cost to society of reducing marriage, that is, the costs of federal, state, and local governmental programs necessary to offset negative societal consequences. For example, studies have shown that children in single-parent homes have a higher probability for negative social outcomes such as under-education, delinquency, and incarceration (Demuth and Brown 2004; Olshewsky et al. 2001; Smith et al. 1999). Consequently, it may be argued that this places an increased burden on society to support public and private organizations that maintain related rehabilitation programs and infrastructure. Research by Alm and Whittington (1995, 1996, 1999) and Sjoquist and Walker (1995) suggests that reduction or elimination of MPT may assist in changing human behavior.

Researchers posit that public policy can alter human behavior. Unfavorable tax consequences of marriage were associated with divorce and lower marriage rates (Smith et al. 1999; Keely 1987). A theory underlying these findings is that the extra tax burden is a financial stress and, such causes marital difficulties that ultimately may contribute to divorce. Again, the damaging effect of MPT has a greater impact on lower income families. MPT not only discourages contemplated marriages, but also may present a danger to existing unions.

Prior research confirms that MPT is disruptive to marital stability and society. Extensive research indicates the need of children for stability. When home life is destabilized, a significant decline in child well-being indicators such as school performance and behavior problems occurrs (USA Today 2005). Children in two-parent homes are two to three times less likely to have emotional problems, and boys are less likely to participate in crime (Colson 2005). The link between home environment and crime is well documented (Demuth and Brown 2004; Morse 2003; Olshewsky et al. 2001; Jackson 1997). Between 67 and 75 percent of all prison inmates were not raised in a two-parent home. In certain juvenile correction systems, for example Wisconsin, the number is approximately 90 percent (Colson 2004).

A consequence of MPT is that it may discourage marriage (Alm and Whittington 1995, 1996, 1999). If the unmarried taxpayer becomes a single-parent, this may lead to greater challenges in child-rearing. Compounding this problem is the assertion that the tax code provides financial incentives to taxpayers to have and raise children outside of marriage. A favorable filing status, child credit, and a higher scaling of the EIC, are all available to single taxpayers. As such, the current tax system may encourage a single-parent family or unmarried two-parent family, impacting on adults, children, and society. Reducing the marriage rate may be an unintended consequence of our current tax system. Elimination or at least minimization of MPT would be a step toward correcting this unintended consequence.

If we examine the nature of the earnings distribution within families to determine which couples specifically incur MPTs, we note instances where married couples pay higher taxes than unmarried couples, even when the income levels of the couples are the same. In families where married spouses have greater differences between the earnings of husband and wife, the MPT is less than in instances where the husband and wife earn similar amounts of income.

RESEARCH DESIGN AND HYPOTHESIS DEVELOPMENT

The objectives of this paper include determining whether a shift in the amount and the distribution of MPT (MBT) occurs between the tax laws in effect prior to EGTRRA 2001 and the changes enacted by JGTRRA 2003. By obtaining population data from income tax returns filed for the year 2000, directional sign predictions are made based on the changes expected using the criteria from the most recent tax act. In terms of estimating the income-split between husband and wife, the methodology used is similar to that employed by Feenberg and Rosen (1995). The key difference between this and earlier studies, including Feenberg and Rosen (1995), is that aggregate population data, rather than sample data is used. A decrease in the amount and distribution of MPT is anticipated as a result.

In addition to the changes enacted in JGTRRA 2003, a shift in the proportion of two-wage earning families also affects the distribution of MPT. Population data of individual tax returns filed for 2000 is available from the Statistics of Income (SOI) Division of the IRS. Using this data, the analysis is useful in approximating and quantifying MPT under the pre-2001 and 2003 tax acts and stratifying the effect across the various tax rate brackets of married filing jointly taxpayers. Consistent with prior research, this study uses actual income tax return data. Population data will also be obtained from the 2000 Census Bureau Current Population Survey (CPS) to evaluate the impact of the tax law changes on the MPT to CPS earnings.

The research questions are as follows, as stated in the alternate form:

RQ1: What is the change between the net effect of the MPT in year 2000 and the projected net effect of the MPT as the result of JGTRRA 2003?

Congress specifically intended to focus on a substantial reduction in MPT. Thus, Research Question 1 posits that there should be empirical evidence of a decreasing shift or complete elimination in MPT situations resulting from the new tax act.

RQ2: What is the change between the distribution of MPT in year 2000 and the projected distribution of MPT throughout all classes of jointly filed income tax returns as the result of JGTRRA 2003?

Research Question 2 addresses the distributional effects between income brackets and assesses the fairness (e.g. horizontal equity) of the reduction in MPT. If Congress truly intended to maintain horizontal equity among various classes of taxpayers, there should be evidence of parity between taxpayers of similar classes of income.

METHODOLOGY

As described earlier, this study utilizes population data from the SOI Division of the IRS and the 2000 CPS data. SOI data consists of tax returns filed by married filing joint taxpayers for the year 2000. Base year tax liabilities are calculated and verified using the tax law in effect for 2000 and then compared to the tax liabilities as re-calculated under JGTRRA 2003. Then, consistent with prior research, the income, deductions, and dependents of the married couple are split according to various assumptions and the tax liabilities of the couple are recalculated as that of two separate single individuals.

Since SOI data does not contain specific information as to the division of income and expenses between spouses, CPS data is employed to allocate consistently the various items of income, expenses, dependency deductions, credits, and other items. CPS data is provided on a monthly basis and is derived from a collaboration of data between the Bureau of Labor Statistics and the Bureau of the Census. CPS data contains employment, earnings, and demographic data. Use of CPS data is important to this study since, when splitting the income and deductions between spouses, certain assumptions must be made. Allocation assumptions in earlier studies divided various individual tax items via a strategy whereby the dependents and deductions were awarded to the spouse with the higher income. This study allocates dependency exemptions and deductions in accordance with the symmetry found in the CPS data.

DATA AND ANALYSIS

In a study of education, employment, health costs, and crime levels (socioeconomic status), Olshewsky et al. (2001) approximate, through a series of extrapolations, that the elimination of MPT could reduce the number of single-parent families by as much as 36%. Hypothetically, the government would give up $33 billion in tax revenues from the MPT but may save even more in curtailment of direct and indirect social program expenditures (e.g. drug rehabilitation and prison facilities). Of greater importance is the impact on the people, arguably a happier, better adjusted society as a whole.

To maintain integrity of the findings and extrapolate the effect of net MPT to the population, income levels representative of comparable income tax brackets for pre- and post-JGTRRA were combined. Using SOI data, Yau, et al. (2004) chose midpoints of the income intervals to represent the average taxpayer for that group. This data also provides the number of men and women who filed joint income tax returns, the corresponding gross amount of income reported for each group, and thus, average salaries and wages for men and women taxpayers were calculated for each interval. Exhibit 3 presents an extrapolation of the data by income interval, number of men and women taxpayers, total income, average income, and provides a percentage mix of the earnings between spouses in jointly filed income tax returns.

The distributional effect between men and women was derived by hypothetically "marrying" the average male filing a joint income tax return with the average female who also filed a joint income tax return for each interval and subsequently calculating the proportionate share of earnings for the couple in each interval. The measurement of this distributional 'mix' allows the calculation of MPT for pre- and post-JGTRRA; therefore, our results are comparable to prior research. The net effect of the change in the tax law is extrapolated to the population as a whole, an improvement over prior research that used samples of tax returns filed rather than population data for all returns filed. Consistent with prior research, calculations were performed at the various marginal income tax brackets for pre-JGTRRA in order to determine if a net MPT (MBT) or benefit exists within each income interval, and overall. To calculate the income for those individuals who file on a single basis, income was split between the couples in each interval using the mix derived from the preceding distribution of wage information. Exhibit 4 reports a summary of MPT in each bracket for pre-JGTRRA for a couple filing married filing joint versus single (separately), with zero dependents.

SOI data does not contain specific information as to the division of income and expenses between spouses, thus various assumptions are necessary to assess the magnitude of MPT. Initially, our tests were conducted using data from the 2000 Supplementary Survey Profile of the U.S. Census Bureau. Median earnings for male full-time, year-round workers was calculated as $38,244 (57% of the two-wage earner couple's joint income) and median earnings for female full-time, year-round workers was estimated as $28,720 (43% of the two-wage earner couple's joint income). Also, the Census survey estimated that approximately 52% of all married couples have both spouses in the labor force. Using the 57% husband/43% wife ratio to re-allocate total income between the couples in the single tax calculations to calculate MPT within each tax rate bracket is too simplistic and does not provide a meaningful generalization to the population. The income division percentages within each income interval are used to calculate net MPT.

Exhibit 5 reports the results of MPT for married filing joint couples with zero dependents using 2003 tax rates. The findings indicate that MPT, while substantially reduced, is still very much evident in lower and higher marginal tax rate brackets.

For situations in which couples in two-income married-couple households have more evenly split incomes than on average, substantially higher MPT will be incurred. For example, the average income split for the married couple with combined wages of $37,500 is 63.7 percent for the husband and 36.3 percent for the wife (IRS 2004). However, married couples with $37,500 of combined wages who have a 50-50 income split and would incur an even higher MPT.

Testing of Research Question 1 indicates a significant change (decrease) in overall MPT as the result of JGTRRA 2003, as shown by the decrease in overall average MPT. The change was from an overall average MPT of $311 (Exhibit 4) to an overall average MBT of $11(Exhibit 5). Testing of Research Question 2 reveals that while there has been an overall reduction in MPT, significant disparity still exists among various classes of income. Differences between MPT for each of the marginal tax rate brackets for pre- vs. post-JGTRRA reveal that, on average, net MPT has decreased, resulting in an overall net MBT in middle income tax rate brackets. On the other hand, MPT was moderately reduced in lower and upper brackets, and the average couple in these brackets still incurs MPT. Furthermore, depending on their income split, some married couples in all income brackets still incur MPT. The closer the income split is to 50-50, the greater MPT.

IMPLICATIONS OF PROVIDING INCENTIVES FOR MARRIAGE

Much tax policy research investigates the application of horizontal equity and its application across various income levels. Conceptually, horizontal equity refers to perceived fairness whereby 'those with equal status or income should be treated equally.' Hence, as it applies to taxes, taxpayers in similar classes of income should pay the same tax and enjoy the same benefits. These benefits conceivably are in the form of goods and services received from the government. Conceptually, a fair or equitable tax policy is one in which the combined income of a married couple would incur the same tax liability as the combined income of two single individuals. However, this is not as simple as it may appear, as calculations are complicated by other factors (such as different tax rate brackets, dependents, and the EITC).

The premise of a government's duty of fiduciary care to its taxpayers and its responsibility to provide its citizens with the optimum utilization of collected revenues and minimization of overall costs raises several issues of concern. In the case of MPT, the impact of the perceived fairness of MPT in providing incentives or disincentives to preserving a two-parent married-couple family unit is one issue. Also related to this issue is whether or not it is fair to allow married individuals to receive MBT, by allowing them to be taxed as if each earned exactly half of their combined income (even when one taxpayer earned all the couple's combined income). This can be regarded as a question of social justice.

A primary justification for eliminating tax disincentives and providing tax incentives that encourage married-couple families is that married couples provide a better environment for raising children, our next generation of taxpayers. Social research describes the benefits of two-parent married couple families (e.g. Demuth and Brown 2004, Morse 2003, Deleire and Kalil 2002, McLanahan 1996). On one side are single taxpayers, who may favor equal taxation of income of all individual taxpayers, married or not. Applying this view would eliminate the married filing jointly option, eliminating not only MPT but also the benefits of taxing combined, and equally apportioned, income associated with the married filing jointly option.

The needs and rights of children must also be considered. Evidence of human history, natural law, and current studies all reveal that the role of spouses in the family is complimentary. Children raised in a two-parent home are much more likely to avoid negative social outcomes such as under-education (McLanahan 1996; Dawson 1991), violent crimes (Parker and Johns 2002), substance abuse (Deleire and Kalil 2002; Hoffman and Johnson 1998), incarceration (Jackson 1997; Morse 2003), and illegitimate births (Demuth and Brown 2004). Seton Hall University law professor John Coverdale (2007, p. 7) summarizes research on family structure as follows: "[C]hildren raised in intact nuclear families have better outcomes on virtually every index than children raised in any other setting. This is demonstrated by numerous studies based on large national samples over extended time frames that control for other variables such as parents' race, education, and income level."

UNDER-EDUCATION

McLanahan (1996) states that children in single-parent families have poorer school attendance records and lower grades than children in two-parent families. School dropout rates for children in single-parent families are twice that of those in two-parent families. Further, children from single-parent homes are less likely to graduate from college and more likely to become single-parents themselves. McLanahan, a single parent herself, developed an interest in this research in an attempt to support her conviction that single parents do "just as good a job of raising children as married moms." She concluded, however, that evidence overwhelmingly indicates that, on average, children do much better in two-parent families.

Research has also demonstrated that children from broken homes perform more poorly in reading, spelling, and math, are more likely to repeat a grade, have higher dropout rates, and lower rates of college graduation. As compared to children from two-parent families, those from single-parent or broken homes exhibit more conduct problems and symptoms of psychological maladjustment, lower academic achievement, social difficulties, and poorer self-concepts (Massachusetts 2004). Dawson (1991), in her review of the National Health Interview Survey of Child Health, reported that "children from disrupted marriages were over seventy percent more likely than those living with both biological parents to have been expelled or suspended." Children from broken homes are twice as likely as those from intact families to drop out of school and more likely to exhibit more health, behavioral, and emotional problems, become involved in crime and drug abuse, and have higher rates of suicide (Zill et al 1993).

VIOLENT CRIMES

Researchers have studied the relationship between family structure and delinquency. In 1925, researchers interested in juvenile delinquency found that almost twice the number of institutionalized (or delinquent) youths came from broken homes (Burt 1925). Criminology literature reveals that family is an essential source of community, providing stability and supervision to form a barrier against violence and that children from broken homes are more delinquent than those from intact families. Parker and Johns (2002) posit that family is a mechanism of social control and an essential element in reducing crime (specifically homicide). They identify constructs of social control occurring in cases of family disruption (i.e., divorce and single-parent households).

SUBSTANCE ABUSE

In addition to experiencing lower educational achievement and higher involvement in crime, studies have found that children from broken or single-parent homes are more likely to engage in drug and alcohol abuse. Using data from the National Educational Longitudinal Study, Deleire and Kalil (2002) observed that teenagers from single-parent families were more likely to initiate smoking or drinking, and initiate sexual activity, but less likely to graduate from high school or to attend college than those from two-parent families. Hoffman and Johnson (1998) concluded that children in traditional two-parent home environments engaged in fewer harmful behaviors such as, delinquency, substance abuse, premarital sex, and suicide.

INCARCERATION

Research indicates that incarceration of one (or both) parents has particularly detrimental consequences to the future development of children. Colson (2004) states that more than half of prison inmates are from broken families. Role models (father and mother) are moral influences that inform the child's consciences early in life; thus, any influence that causes family breakdown is likely to increase social disorder and crime.

In Parents or Prisons, Morse (2003) speculates that for some, prisons are substitutes for parents but, obviously, poor substitutes for caring, nurturing parents. She states that without two parents, a child is more likely to end up in the criminal justice system at some point, and the prison will perform the parental function of supervising and controlling behavior. Basic development of self-control, reciprocity, conscience, trust, and empathy takes place in the family units. Morse comments that a free society must have people with consciences. The great majority of people obey laws voluntarily. If citizens fail to conform to the law, they must be compelled to do so or the public protected when they do not. Thus, lacking two parents, a typical child has greater difficulty learning the moral basics. Colson (2006) indicates that failure to teach moral basics to children is a key reason for the ten-fold increase in the U.S. prison population (from 230,000 in 1976 to 2.3 million in 2006--a period in which the U.S. population increased by about 50 percent). The U.S. now leads the world in per capita prison incarceration.

Morse (2003) provides an analysis of the estimated cost of incarceration to society. Using California as an example, the Department of Corrections allocation accounted for approximately six percent ($5.2 billion) of the entire state budget in 2002-2003. On a per capita basis, this was approximately $26,700 per adult inmate per year and $49,200 per juvenile offender per year. She poses an interesting question: What if minors in the juvenile system were functioning well enough that they could be a normal part of society and thus a part of the educational system rather than the juvenile delinquency program--how much would this save the taxpayers?

Morse (2003) extrapolates that taxpayers pay about $8,568 per year per student K-12, $4,376 per year per student at a community college and $17,392 per year per student in the University of California system, concluding that exchanging school for prison yields tremendous savings for society. Of course, society receives many additional benefits from reduced crime, such as the tangible economic savings from reduced theft and robbery. In addition, they are numerous intangible benefits, such as people no longer living in fear of becoming a victim of crime, as well as the benefit to those who live a productive rather than a criminal life.

ILLEGITIMATE BIRTHS

More than half of all children will spend at least some time in a single-parent family. Using data from the 1995 National Longitudinal Survey of Adolescent Health, Demuth and Brown (2004) estimate that nearly a third of all children are born to unmarried mothers. Constructs through which living with a single parent increases delinquency and indicate that adolescents in single-parent families are more likely to engage in more serious forms of harmful behavior.

Then-Vice-President Dan Quayle sparked controversy when he criticized the television series Murphy Brown depicting a child born out of wedlock (CNN 2002). The trials and tribulations of a single-parent female role model were portrayed and controversy regarding traditional family values and concerns of raising a fatherless child were prompted. McLanahan (1994) found that children raised by a single parent are disadvantaged across a broad array of outcomes. These children are twice as likely to drop out of high school, two and a half times more likely to become teen mothers themselves, and one hundred-forty percent more likely to be idle (out of school and out of work) as are children who were raised by two parents. She adds that children from single parent families also have poorer school attendance records, lower grades, lower college aspirations, and, as adults, higher rates of divorce.

McLanahan (1996) points out that out-of-wedlock births have increased in the U.S. since the 1940's and suggests that more should be done to assist parents with the costs of raising children. She emphatically states that a better way to encourage marriage is to make certain that parents are not penalized for marriage, but our current system of income transfers and taxation does just that. The issue that must be addressed is why MPT continues to be imposed.

Many believe that eliminating MPT, and allowing a benefit (i.e., taxing a married couple's combined incomes apportioned equally under the married filing jointly option) results in social justice. Eliminating MPT strengthens the married two-parent family and advances society's ability to raise future generations of citizens. An unmarried cohabiting couple also provides children with two parents, but statistics indicate that half of cohabiters break up in five years, while only 15 percent of married couples separate during this period (USA Today 2005).

Estimated costs to society associated with the criminal justice system, prison facilities, substance abuse treatment, lost productivity, and under-education attributable to single-parent households have been estimated to exceed $300 billion (Olshewsky et al. 2001). Eliminating MPT is expected to strengthen two-parent households and thereby reduce the number of children in single-parent households.

Overwhelmingly, research indicates the benefit to children of two-parent married-couple households over single-parent households. In addition, the benefits of marriage extend beyond the well-being of children. Warren (2003) indicates that married persons are better off emotionally, physically, financially, and vocationally than are unmarried partners. Adults benefit from marriage, regardless of whether or not they are parents. Essentially, strengthening marriage, by eliminating MPT, is beneficial to children and adults.

LIMITATIONS, CONCLUSIONS AND FUTURE RESEARCH

While researchers calculate and measure the impact of a change in tax law on MPT among various classes of taxpayers, it is difficult to ascertain the exact impact of MPT on marriage. Further, division of income and deduction items found in this body of research is based on various assumptions and, as plausible as these assumptions may seem, they are nevertheless estimates of what could occur, not necessarily what does occur. Additionally, in attempting to estimate the overall true cost to society as the result of tax changes, it is difficult to precisely extrapolate those results to the population as a whole.

Many regard the marriage penalty tax as contradictory to American values and general principles of fairness. Survey data and anecdotal evidence indicate that most people believe MPT should be eliminated. Results of this study indicate that while a significant reduction in MPT has occurred, substantial disparity still remains throughout various class of income. Taxpayers in the lowest and highest marginal income tax rate brackets received smaller benefit and taxpayers in the middle marginal income tax rate brackets received greater proportional benefit. Congress has voiced publicly its intention to substantially reduce the inequity of MPT. Perhaps Congress can do so on a 'net' basis but so far has been unable to accomplish this objective throughout all classes of income. This study confirms that MPT continues to exist for many married taxpayers. To make matters worse, current reductions in MPT will 'sunset' or expire after 2010, unless Congress votes to extend provisions of JGTRRA.

The family is widely regarded as the foundation of our civilization (Hagelin 2007, Colson and Fickett 2005; Colson 2003; Olshewsky et al. 2001, Smith et al. 1999). Considering the positive aspects of marriage on society--stable family structure, emotional health of children, and physical health of adults (with or without children), placing a higher tax burden on married couples appears contrary to the best interests of the nation. Tax researchers may posit that while there are some couples who pay MPT, this is offset by other couples that receive MBT. While this may be true, it is little consolation to the marriages negatively affected by the MPT. Based on this research, one can infer that the MPT via Congress is harming marriage and the best interests of society, both children and adults. The future of American civilization will largely depend on the vitality of marriage and the family. The MPT is detrimental to both. Families are the foundation of civilization; when families fail, civilization collapses. In the final analysis, the MPT is anti-family and thereby anti-civilization.

Perhaps this study will stimulate further interest in determining the impact of changes in tax policy on behavior and the public and social costs associated with MPT. Our work may be extended by utilizing various assumptions for re-allocation of tax variables including income, expense, and dependents. Division of these items may affect the impact of changes in tax law. Calculations involving the separation of the taxpayers from married to single/head of household may involve assigning the dependents to the highest wage earner, or to the wife, or splitting them equally among separating ex-spouses. Similarly, income and expense items could be re-allocated among separated taxpayers.

Ultimately, if Congress wishes to maintain revenue neutrality, alternate sources of revenue must be considered. Researchers might investigate other potential sources of tax revenue to replace that lost by the minimization or elimination of MPT.

DEDICATION

Researcher, scholar, and marriage penalty tax expert, the late Leslie A. Whittington, Professor of Public Policy at Georgetown University in Washington D.C., was killed on 9-11, 2001 when American Airlines Flight 77 crashed into the Pentagon with 59 people aboard. Professor Whittington, her husband, Charles Falkenberg, and their two daughters, 8-year old Zoe and 3-year old Dana were traveling to Los Angeles and then on to Australia where Whittington had been named a visiting fellow at Australian National University in Canberra.

ACKNOWLEDGMENTS

The authors wish to thank the reviewers, discussant, and participants at the American Accounting Association Annual Meeting and research workshop participants at Texas A&M University.

REFERENCES

Allen, J. 2008. Newly Elected House Democrats Characterize Budget as Tax Hike. The Hill, Website: TheHill.com (June 9).

Alm, J., and L.A. Whittington. 1995. Does the income tax affect marital decisions? National Tax Journal 48 (4): 565-572.

Alm, J., and L.A. Whittington. 1996. The rise and fall and rise ... of the marriage tax. National Tax Journal 49 (4): 571-589.

Alm, J., and L.A. Whittington. 1999. For love or money? The impact of income taxes on marriage. Economica. 66 (263): 297-316.

Brozovsky, J., and A.J. Cataldo. 1994. A historical analysis of the marriage penalty tax. Accounting Historians Journal 21 (1): 163-174

Bull, N., J. Holtzblatt, J.R. Nunns, and R. Rebelein. 1999. Defining and measuring marriage penalties and bonuses. Office of Tax Analysis (OTA), U.S. Department of Treasury. OTA Paper 82-Revised.

Burt, C. 1925. The young delinquent. London: University of London Press.

Census Bureau. 2000. Current Population Survey. Washington, D.C., U.S. Census Bureau.

CNN. 2000. Senate Poised to Vote on Marriage Penalty Tax Cut. CNN, website: CNN.com (July 17).

CNN Access. 2002. Quayle: Ten Years After Murphy Brown. CNN, Website: http://www.cnn.com/2002/allpolitics/05/09/quayle.cnna/ (May 10).

Colson, C. 2003. The new normal: Singleness in America. Breakpoint No. 031211: December 11.

Colson, C. 2004. Take your choice: Parents or prisons. Breakpoint No. 040624: June 24.

Colson, C. 2005. Life and options in suburbia: Desperate housewives. Breakpoint No. 050201: February 1.

Colson, C. and H. Fickett. 2005. The good life: Seeking purpose, meaning and truth in your life. Tyndale.

Colson, C. 2006. The Moussaoui Verdict, Freud Is Alive and Well. Breakpoint No. 060509: May 9.

Cook, E.D., L.M. Smith, and L.C. Smith. 2001. A delayed fix for the marriage penalty. Taxes, the Tax Magazine 79 (10): 15-22.

Cook, Robin. 2006. Congressional Democrats May Shift Tax Policy, Say Yin, Doran. Virginia Law News. Website: http://www.law.virginia.edu/html/news/newsevents.htm (November 16).

Coverdale, John. 2007. Letter to Time Magazine. Family News from Dr. James Dobson (January): 7.

Dawson, D. 1991. Family structure and children's health and well-being: Data from the 1988 national health interview survey on child health. Journal of Marriage and the Family 53: 573-584.

Deleire, T., and A. Kalil. 2002. Good things come in threes: Single-parent multigenerational family structure and adolescent adjustment. Demography 39 (2): 393-414.

Demuth, S., and S. Brown. 2004. Family structure, family process, and adolescent delinquency: The significance of parental absence versus parental gender. Journal of Research in Crime and Delinquency 41 (1): 58-81.

Economic Growth and Tax Relief Reconciliation Act of 2001. 2001. Washington, D.C., U.S. Congress.

Feenberg, D.R., and H.S. Rosen. 1995. Recent developments in the marriage tax. National Tax Journal 48 (1): 91-101.

Gelardi, A. M. 1996. The influence of tax law changes on the timing of marriages: A two-country analysis. National Tax Journal 49 (1): 17-30.

Hagelin, Rebecca. 2007. Endangered: Western Civilization. Press and Media, The Heritage Foundation Website: heritage.org (March 9).

Hoffman, J. P., and R. A. Johnson. 1998. A national portrait of family structure and adolescent drug use. Journal of Marriage and the Family 60 (August): 633-645.

Holtzblatt, J., and R. Rebelein. 2000. Measuring the effect of the EITC on marriage penalties and bonuses. National Tax Journal 53 (4): 1107-1133.

Internal Revenue Service (IRS). 2004. Statistics of Income (SOI). Washington, D.C., Internal Revenue Service.

Jackson, K. 1997. Differences in the background and criminal justice characteristics of young black, white, and hispanic male federal prison inmates. Journal of Black Studies, 27 (March): 494-509.

Jobs and Growth Tax Relief Reconciliation Act of 2003. 2003. Washington, D.C., U.S. Congress.

Keely, M. 1987. The effects of experimental negative Income tax programs on marital dissolution: Evidence from the seattle and denver income maintenance experiments. International Economic Review (February): 241-257.

Kinney, William R. 2003. Accounting Research Workshop Presentation, Texas A&M University (July 10).

Massachusetts Family Institute. 2004. Why marriage matters. Issues in Focus. http://www.mafamily.org (June 30).

McIntyre, R.S., and M.J. McIntyre. 1999. Fixing the "marriage penalty" problem. Valparaiso L. Review 33: 907-946.

McLanahan, S. 1994. The consequences of single motherhood. The American Prospect 5 (18): June 23.

McLanahan, S. 1996. Double the trouble. Psychology Today 29 (3): 18.

Moffit, R., R. Reville, and A. Winkler. 1998. Beyond single mothers: Cohabitation and marriage in the AFDC program. Demography 35 (3): 259-278.

Morse, J. R. 2001. Love and Economics. Spence.

Morse, J. R. 2003. Parents or prisons. Policy Review 120 (August-September): 1-14.

Olshewsky, Steve, L.M. Smith, and L.C. Smith, Jr. 2001. Accounting for the economic costs of the marriage penalty tax. Tax Notes 90 (8): 1093-94.

Parker, K. F., and T. Johns. 2002. Urban disadvantage and types of race-specific homicide: Assessing the diversity in family structures in urban context. Journal of Research in Crime and Delinquency 39 (3): 277-303.

PBS. 2000. Marriage Penalty. PBS, Website: Pbs.org (February 10).

Pew Research Center. 2005. Beyond Red vs. Blue. Pew Research Center for the People & the Press, Website: PeoplePress.org(May 10).

Popenoe, David and Barbara D. Whitehead. 2000. The state of our unions: The social health of marriage in America, National Marriage Project, Rutgers University (June).

Sjoquist, D.L., and M.B. Walker. 1995. The marriage tax and the rate and timing of marriage. National Tax Journal 48 (4): 547-558.

Smith, L.M., L.C. Smith, and E.D. Cook. 1999. Toward family-friendly public policies. Southwestern Economic Review 26 (1): 137-154.

Stevenson, R.W. 2000. Republicans Offer Tax Cut For Couples. New York Times, Website: nytimes.com (February 1).

USA Today. 2005. The costs children pay. USA Today (September 22): 10A.

U.S. Congressional Budget Office. 1997. For better or worse: Marriage and the federal income tax. Washington, D.C., June.

The White House. 2003. Excerpts from Economic Security press release. Remarks by President Bush at signing of Jobs and Growth Tax Relief Reconciliation Act of 2003, May 28.

Warren, N.C. 2003. The cohabitation epidemic. Focus on the Family June-July: 10-12.

Whittington, L.A., and J. Alm. 2001. Tax reductions, tax changes, and the marriage penalty. National Tax Journal 54 (3): 455-472.

Wiggins, C.E., Jr., D. L. Crumbley, and N. G. Apostolou. 1986. A critical analysis of the marriage tax penalty. The Journal of the American Taxation Association 7 (2): 60-75.

Yau, E., K. Gurka, and P. Sailer. 2004. Comparing salaries and wages of women shown on forms W-2 to those of men, 1969-1999. Statistics of Income Division, Internal Revenue Service, January.

Zill, N., D. Morrison, and M. Coiro. 1993. Long-term effects of parental divorce on parent-child relationships, adjustment, and achievement in young adulthood. Journal of Family Psychology 7(1): 91-103.

Frederick J. Feucht, Prairie View A&M University

L. Murphy Smith, Texas A&M University

Robert H. Strawser, Texas A&M University

Exhibit 1

Standard Deductions for Tax Years Pre- and Post-JGTRRA

 MF-Joint MF-Sep

Pre-JGTRRA
2000 7,350 (167% of Single) 3,675
Post-JGTRRA
2003 9,500 (200% of Single) 4,750 (50% of MFJ=S)
2004 200%
2005 174%
2006 184%
2007 187%
2008 190%
2009 200%
2010 200%

 Single Head of
 Household
Pre-JGTRRA
2000 4,400 6,450
Post-JGTRRA
2003 4,750 7,000
2004
2005
2006
2007
2008
2009
2010

Exhibit 2: Net Marriage Penalty Tax Based on Prior Research

Study / year published Tax Average Penalty
[tax act] Year % and $

OTA 1999 1999 48 and $1,141

CBO 1997 1996 42 and $1,380

CBO 1997 1996 47 and $1,750
(all itemized)

Alm and 1994
Whittington 1996

Feeberg and Rosen 1994 51 and $1,244
1995 [OBRA 93]

Feeberg and Rosen 1993 51 and $898
1995

Rosen 1987 1988 40 and $1,091
[TRA 86]

Rosen 1987 1986

Alm and 1980
Whittington 1996

Study / year published Average Benefit Net
[tax act] % and $ Penalty

OTA 1999 41 and $1,274 31

CBO 1997 51 and $1,300 --

CBO 1997 49 and $1,350 --
(all itemized)

Alm and $375 *
Whittington 1996

Feeberg and Rosen 38 and $1,399 $124 *
1995 [OBRA 93]

Feeberg and Rosen 38 and $1,577 $(143) *
1995

Rosen 1987 53 and $609 $119 **
[TRA 86]

Rosen 1987 $529 **

Alm and $300 *
Whittington 1996

* 1994 dollars and ** 1988 dollars

Exhibit 3: Distribution of Wages between Husband and Wife
in Dual Wage-Earner Families

Men joint filers

Income Number of Total $ * Average
Interval taxpayers (in $1,000's) Salaries
by Size and Wages
$ of AGI *** $

Under 5K ** 269,185 3,625,359 13,468
10,000 1,400,760 11,073,920 7,906
20,000 2,725,002 39,239,895 14,400
37,500 9,892,261 255,574,311 25,836
75,000 15,668,855 681,871,934 43,518
150,000 5,197,754 407,510,131 78,401
300,000 854,387 146,544,803 171,520
600,000 657,630 156,859,555 238,523
1,000K or more ** 136,077 151,594,427 1,114,033
Overall averages 36,801,910 1,853,894,334 50,375
 (all intervals)

Income Number of Total $ * Average
Interval taxpayers (in $1,000's) Salaries
by Size and Wages
$ of AGI *** $

Under 5K ** 276,000 2,116,731 7,669
10,000 1,208,996 7,768,129 6,425
20,000 2,266,236 22,613,860 9,979
37,500 8,247,782 121,485,004 14,729
75,000 13,868,379 349,627,413 25,210
150,000 4,359,367 172,925,267 39,668
300,000 571,900 33,817,488 59,132
600,000 409,501 29,402,946 70,923
1,000K or more ** 57,928 14,424,850 249,015
Overall averages 31,266,089 753,821,689 24,110
 (all intervals)

Income Distribution of avg earnings
Interval in 2 wage-earner MFJ households
by Size
$ of AGI ***

 Husb Wife

Under 5K ** 63.7% 36.3%
10,000 55.2% 44.8%
20,000 59.1% 40.9%
37,500 63.7% 36.3%
75,000 63.3% 36.7%
150,000 66.4% 33.6%
300,000 74.4% 25.6%
600,000 77.1% 22.9%
1,000K or more ** 81.7% 18.3%
Overall averages 67.6% 32.4%
 (all intervals)

* Monetary amounts are in thousands of dollars, averages are in
whole dollars
** Disregarded interval due to unknown variance (range from $5,000
to negative and above $1,000,000 are indeterminate).
*** Source: Data derived from "Comparing Salaries and Wages of Women
Shown on Forms W-2 to Those of Men, 1969-1999", p.283, IRS SOI
Jan. 2004

Exhibit 4: Pre-JGTRRA Impact of MPT (MBT) within classes of income
by Marginal Tax Bracket, No Dependents
Married Filing Jointly (MFJ) vs Separately (as Single)

Pre-JGTRRA Combined Spouse 1 Spouse 2
(2000) Wages Husband Wife
Marg.Tax (% split) * (% split) *
Brack. % $ split $ split

 (55.2%) (44.8%)
15% $10,000 5,520 4,480
 (59.1%) (40.9%)
15% $20,000 11,820 8,180
 (63.7%) (36.3%)
15% $37,500 23,888 13,612
 (63.3%) (36.7%)
28% $75,000 47,475 27,525
 (66.4%) (33.6%)
31% $150,000 99,600 50,400
 (74.4%) (25.6%)
36% $300,000 223,200 76,800
 (77.1%) (22.9%)
39.6% $600,000 462,600 137,400
 (67.6%) (32.4%)
Overall avg. $74,485 50,375 24,110

Pre-JGTRRA Tax liab. MPT or (MBT)
(2000) M F J M F J vs Single
Marg.Tax Tax due Assume both taxpayers file
Brack. % (refund) Single + 0 dep

15% $ (27) ** $ 668 **
15% 1,061 388
15% 3,686 221
28% 11,681 767
31% 33,607 1,583
36% 88,247 5,405
39.6% 207,357 13,127
Overall avg. 11,527 311

* Source: Data derived from "Comparing Salaries and wages of Women
Shown on Forms W-2 to Those of Men, 1969 - 1999", p.283, IRS, SOI,
Jan 2004.

** Due to Earned Income Tax Credit (EITC).

Exhibit 5 : Post-JGTRRA Impact of MPT (MBT) within classes of income
by Marginal Tax Bracket, No Dependents
Married Filing Jointly (MFJ) vs Separately (as Single)

Post-JGTRRA Combined Spouse 1 Spouse 2
(2003) Wages Husband Wife
Marg. Tax (% split) * (% split) *
Brack. % $ split $ split

 (55.2%) (44.8%)
10% $10,000 5,520 4,480
 (59.1%) (40.9%)
10% $20,000 11,820 8,180
 (63.7%) (36.3%)
15% $37,500 23,888 13,612
 (63.3%) (36.7%)
25% $75,000 47,475 27,525
 (66.4%) (33.6%)
28% $150,000 99,600 50,400
 (74.4%) (25.6%)
33% $300,000 223,200 76,800
 (77.1%) (22.9%)
35% $600,000 462,600 137,400
 (67.6%) (32.4%)
Overall avg. $74,485 50,375 24,110

Post-JGTRRA Tax liab. MPT or (MBT)
(2003) M F J M F J vs Single
Marg. Tax Tax due Assume both taxpayers file
Brack. % (refund) Single + 0 dep

10% $ (169) ** $ 555 **
10% 443 235
15% 2,589 (55)
25% 8,476 (862)
28% 27,813 (110)
33% 76,787 3,377
35% 181,882 9,268
Overall avg. 8,339 (11)

* Source: Data derived from "Comparing Salaries and wages of Women
Shown on Forms W-2 to Those of Men, 1969 - 1999", p.283, IRS, SOI,
Jan 2004.

** Due to Earned Income Tax Credit (EITC).