Fourteen years ago, members of Congress sympathetically listened as divorcees testified to their struggles to raise children while being pursued by the Internal Revenue Service for tax debts, often unknown to them, that were attributable to their ex-husbands’ income. Rather than adopting one of many proposals to end joint and several liability, Congress instead elected to expand the grounds on which these individuals could seek relief from such liability. Since that time, taxpayers have seen a steady expansion of the grounds for so-called “innocent spouse relief” that has evolved through a combination of legislative, administrative, and judicial action. Yet the process for relief remains time-consuming, inefficient, and unpredictable. The majority of initial requests for innocent spouse relief are denied. The taxpayer can appeal administratively and also seek judicial review if relief is denied, but sometimes will spend several years and untold resources in pursuit of a claim that may ultimately be unsuccessful. The process is also a questionable use of Internal Revenue Service personnel, in that it frequently calls upon these employees to address the most intimate aspects of a failed relationship, including spousal abuse, addictions, and mental health problems. These employees often must make a determination based upon a “he-said, she-said” presentation of the facts—an odd task for an agency charged with enforcing the revenue laws.
This article visits the historic rationales for joint and several liability, both in light of the flawed relief process and also in the context of modern-day American society, in which married couples constitute only half of all households and cohabitation is increasingly more common. I conclude that Congress should eliminate the “married” filing statuses and require each married individual to file a separate return. If it did so, joint liability and the innocent spouse relief process would both cease to exist. The historical policy justifications for imposing joint and several liability are no longer rational in light of changed demographics and technological advances. Rather than an “unusual privilege,” which it was long said to be, filing jointly has become a risky conundrum, particularly for low-income taxpayers. As the nation debates tax reform, it is appropriate to rethink the policy of retaining the “married” filing statuses in light of the ways in which family structures, society, and the Internal Revenue Code have changed since joint and several liability was introduced in 1938.
- I. Introduction
- II. How did we get here? (History and purpose of joint filing status)
- A. Shared Risks: Tracing the History of Joint and Several Liability
- B. Examining the Stated Rationales of Joint and Several Liability: Does the Intent Still Make Sense in Twenty-First Century America?
- III. Relief is Available to Some Innocent Spouses, but at What Cost to the System?
- IV. Solutions – Revisiting the Treasury Department’s Recommendation to Expand Innocent Spouse Relief Instead of Eliminate Joint and Several Liability
- A. A Systemic Change: Mandatory Separate Returns for all Fliers
- B. Less Comprehensive Proposals: Proportionate Liability, Allocated Liability, and Adoption of the Divorce Decree
- C. Judging With the Benefit of Hindsight – Did Treasury Make the Right Recommendation in 1998? The Unending Inefficiencies of the Innocent Spouse Process
- V. Neither the code nor america looks like it did in 1938 (or even 1975): why congress should again revisit the call for mandatory separAte filing
- A. A Closer Look at the Changing Demographics
- B. Low-Income Households Are Disproportionately Affected by the Filing Status Options for Married Persons: Why the Code Should Tax the Type of Household, Not the Relationship
- C. A Brief Consideration of How to Restructure the Earned Income Credit in an Individual Filing System
- VI. Conclusion
Married taxpayers have two options when filing a federal income tax return: (1) file a joint return, in which case the taxpayers agree to joint and several liability for any tax due; or (2) file separate returns by electing the “married filing separately” status. It is estimated that joint returns typically comprise ninety-five percent of the income tax returns filed by married taxpayers. The Internal Revenue Code (“Code”) does not favor married taxpayers who file separately; except in very unusual circumstances, these taxpayers will face a higher tax liability by choosing this filing status. Thus, it is almost always in a married couple’s collective financial interest to file a joint return. However, if the couple files jointly and then later divorces, the Internal Revenue Service (“Service”) can pursue collection efforts for the joint years against either or both taxpayers simultaneously, regardless of which spouse’s income created the liability.
One of the simplest and most automated types of collection activity is the refund offset: the Service will apply an income tax refund expected from one year (in whole or in part) to satisfy an outstanding tax liability from a different year. Thus, in many cases involving divorced couples with minor children, the Service will seize the sizable refundable tax credits due to the low-to-middle-income custodial parent in satisfaction of a past debt that, while joint, is wholly attributable to the income of the noncustodial parent. Even if the noncustodial parent is making periodic payments towards the liability, whether voluntarily (under an installment agreement) or involuntarily (through wage garnishment), the refund due to the custodial parent will be automatically seized to further reduce the debt. Thus, the custodial parent may end up paying a significant portion (and in some cases, all) of the joint tax debt, even if the parties had agreed otherwise under the terms of the divorce. This occurs without regard to the parties’ respective income level and ability to pay.
Consider the story of Wendy Bozick, whose joint liability arose from the “traditional” model of the one-earner household: in 2003, her husband Gary earned $460,000 from his law practice. Ms. Bozick, who did not perform significant work outside of the home that year, earned just over $1,000 in 2003. Her husband handled all finances; Ms. Bozick did not have a bank account and her husband did not allow her to use a credit card. Her husband gave her cash to buy groceries for the family. He did not allow her to open the mail. As the filing date for their income taxes approached, Ms. Bozick worried that her husband might not prepare the return. She filed “married filing separately” to report her wages, which she thought she was required to do. Ms. Bozick owed no tax on her separate return due to her low income. When her husband found out, he became furious, telling her that filing separately rather than jointly would cost the family $17,000. This is because a one-earner, high-income family enjoys the benefit of income splitting, which creates a “marriage bonus” as compared to a similarly situated unmarried couple. Recognizing this, her husband had a joint return prepared and had her sign it without giving her a chance to examine it. The joint return, which was filed in December 2004, showed the couple had a tax liability of $137,453. Ms. Bozick petitioned for divorce in July 2005, and Mr. Bozick died in April 2006. Upon his death, the Service collected a portion of the joint liability from his retirement plan, but a balance of more than $100,000 remained on the debt. The Service turned its collection efforts on Ms. Bozick to collect the remainder. By this time, she had joined the work force full-time, but was earning only eight dollars an hour. She was barely supporting herself and her children.
In February 2007, Ms. Bozick filed a request for “innocent spouse relief,” but the Service denied her claim in July of that year, stating, among other reasons, that she failed to show that it would be unfair to hold her jointly liable. In accordance with her rights of judicial review, Ms. Bozick petitioned the United States Tax Court in October 2007 and told her story again at her March 2009 trial. On March 30, 2010, more than three years after her initial request for relief, the court filed its opinion in her favor and relieved her of this joint income tax liability.
This article will examine the problems and inefficiencies that arise from joint and several liability and the innocent spouse relief process. In particular, I will examine the peculiar conundrum faced by low-income taxpayers. Unlike the Bozick family, which received a sizable economic benefit by virtue of the joint filing status, a low-income couple generally will receive only a nominal “marriage bonus” attributable to income splitting. As I will explain, married low-income taxpayers face a conundrum because they cannot receive the earned income credit unless they file a joint return, subjecting themselves to joint and several liability. Meanwhile, if the married couple does file jointly, there is no economic advantage as compared to a similarly situated unmarried and cohabiting couple; in fact, the unmarried couple will typically receive a larger aggregate refund.
The article will focus on the three reasons I believe the joint filing status should be eliminated and all taxpayers should file as individuals: (1) the rationales for joint filing are less applicable than they once were; (2) Congressional attempts to address inequities through innocent spouse relief are a well-intended but fundamentally inefficient solution; and (3) it is necessary to reconsider the policy of retaining the “married” filing statuses in light of the ways in which family structures, society, and the Code have evolved since joint and several liability was enacted in 1938. Congress should reform the current filing-status structures through a lens of simplicity and demographic reality.
The article proceeds from here in four parts. Part II will provide a brief historical context: why and how the U.S. tax system provides different rules and rates for married taxpayers, and the original reasons for imposing joint and several liability on joint returns. Within that context, I will discuss why the traditional rationales underpinning joint and several liability are less convincing today. This is particularly true for low-income taxpayers, as I will discuss.
Part III will present a detailed overview of the procedures for innocent spouse relief. The innocent spouse relief process is time-consuming, inefficient, and results in inconsistent outcomes. It is not uncommon for a taxpayer to spend several years and untold resources trying to obtain the legal relief he or she seeks. The majority of initial requests for innocent spouse relief are denied, meaning the taxpayer must pursue an administrative appeal and often judicial review of the claim. In recent years the process has become ever more liberalized, broadening the grounds for relief and the scope of judicial review if relief is denied. In 1998, when Congress enacted § 6015(f), more taxpayers became eligible for relief and the number of claims increased. Since that time, the Service has regularly received up to fifty thousand requests for innocent spouse relief a year, and it grants fewer than half of them. Taxpayers are statutorily entitled to petition the Tax Court if the Service denies the request for relief; innocent spouse appeals are consistently among the ten most litigated tax issues in federal court each year. The facts set out in the Tax Court’s opinion in Bozick certainly weighed in her favor and portrayed her in a sympathetic light. Yet it took her three years to obtain relief. Her administrative request was denied, and one presumes that the IRS was unwilling or unable to settle the case in her favor after she filed her Tax Court petition. Over the course of three years, imagine how many hours of work must have been devoted to this file: by the IRS Innocent Spouse Unit, the Tax Court, and Ms. Bozick and her attorney. During that time, Ms. Bozick endured financial hardship and struggled to raise her children. While I agree with the Tax Court’s finding in this specific case and support the Congressional intent of the innocent spouse provisions generally, I maintain that the process is burdensome to taxpayers and is an inefficient use of administrative and judicial resources. Moving to an individual filing system would eliminate the need for innocent spouse relief, allowing those resources to be devoted elsewhere.
Part IV examines some of the solutions previously proposed to address the shortcomings of the innocent spouse relief framework. A chorus of scholars, professional associations, and the National Taxpayer Advocate has recommended one solution in particular—to eliminate joint and several liability for joint filers. This solution would address the first two of my concerns; importantly, it would render the innocent spouse relief process unnecessary, because each spouse would only be liable for tax arising from his or her proportionate income. At the direction of Congress, the Department of Treasury studied this and other proposals in 1998 and issued a comprehensive report of its findings. The report concluded that there were sound policy and practical reasons to retain joint filing and joint and several liability for married couples. Instead of adopting the proposals to eliminate joint and several liability, the report recommended modifying the Code to accommodate more cases by expanding the grounds for innocent spouse relief. Congress followed this recommendation, greatly broadening the circumstances under which a joint filer can later apply for relief. As a result, since 1998 there has been a drastic rise in the number of innocent spouse relief claims filed and a corresponding increase in the amount of resources devoted to innocent spouse claims. Part IV examines the reasons the Treasury Department did not believe it was appropriate to replace or reform the joint and several liability standard; with the benefit of hindsight, I evaluate those arguments.
While I believe eliminating joint liability would have been a positive solution, I discuss in Part V why I believe Congress should go even further: it should eliminate joint filing altogether and require all taxpayers to file an individual return of their own income. Part V will elaborate on the third of the concerns I outlined above, namely, that the Code does not properly reflect the demographic reality of twenty-first century America. Today there are many more accepted versions of what constitutes a “household” or “family unit” than there were in the mid-twentieth century, making the “married filing jointly” status an antiquated concept. Marital status should not be used as a proxy for household composition because this does not match the reality. There are many reasons why it is time for the Code to shift away from a system that assumes and favors a married couple with only one spouse in the labor force: the rise of divorced individuals and single-parent households; the rise in the number of two-earner households; the significant increase in the number of couples who live together in a household but do not wed; and the unavailability of this filing status for individuals in same-sex partnerships or marriage.
My argument—that demographic trends matter—will focus in particular on low-income taxpayers. It examines how the value of the earned income credit, a refundable credit primarily benefiting parents that was not added to the Code until several decades after the joint filing status was introduced, has changed the stakes for low-income taxpayers when deciding upon a filing status. Married taxpayers must accept the risks of joint and several liability or else sacrifice eligibility for the refundable earned income credit; choosing the latter may mean the taxpayers forsake thousands of dollars to help support their household.
If the couple files jointly and a deficiency is later determined, it is true that the spouse who did not earn the income or claim the deduction causing the unpaid tax can file an innocent spouse request for relief from joint liability. In fact, low-income taxpayers (as Ms. Bozick was at the time she filed) are especially likely to file for such relief. Taxpayers who request innocent spouse relief are disproportionately low-income women who are unmarried; often they are struggling to maintain a home for children. Most often, it is the ex-wife who requests innocent spouse relief, and she constitutes the lower-earning half of the divorced couple. Low-income taxpayers qualify for pro bono legal representation in their innocent spouse claims, presuming that they know to inquire about such relief and have the time to work with counsel in pursuing it. But it is certainly not an easy, pleasant, or quick process to endure, and there is a high likelihood the taxpayer will not prevail.
Meanwhile, unmarried couples raising children together, a demographic group that has increased at a dramatic rate since 1960, do not face a filing dilemma that might result in having to apply for relief from joint liability. Those low-income divorcées who have endured the innocent spouse relief request process may find it especially galling to learn that an unmarried couple, in fact, can receive a higher overall earned income credit than a married couple with an identical household income.
It would be far more beneficial to married low-income taxpayers if Congress were to amend the Code so as not to condition the earned income tax credit (or any form of perceived “marriage bonus”) on joint filing. Congress can address this conundrum and today’s complex demographic reality without adding further complexity to the Code. Congress should respond to the calls for a simpler and more coherent Code by reducing the number of individual filing statuses from four to two, ignoring marital status while retaining a distinction in the treatment of taxpayers with and without children.
II. How Did We Get Here? (History and Purpose of Joint Filing Status)
Much has been written about the history of joint filing and the rationale behind joint and several liability. In order to appreciate the shortcomings and drawbacks of joint and several liability, it is important to first consider why the current system is as it is.
A. Shared Risks: Tracing the History of Joint and Several Liability
Joint and several liability for married taxpayers has been the settled rule of the Code since 1938, but it was not always the rule. Early revenue laws enacted in 1913 and 1916 imposed taxation on individuals. In 1918, married taxpayers were permitted to file a joint return and could aggregate their income as well as their losses and deductions, but the Code left open the question of how to allocate the liability between spouses.
The idea of joint and several liability originated not with Congress, but in an internal office decision known as I.T. 1575, which was published by the Bureau of Internal Revenue in 1923. Even after the Supreme Court held otherwise, the Commissioner took the position that this type of office decision—published by the Income Tax Unit of the Bureau with a cautionary note that they “have none of the force or effect of Treasury Decisions and do not commit the Department to any interpretation of the law”—evinced the intent of Congress if the underlying revenue acts were reenacted without change. Thus, even though the Code was silent on the issue, in litigation the Commissioner repeatedly advanced the position that joint and several liability was “settled administrative practice.”
In Cole v. Commissioner, the Ninth Circuit Court of Appeals considered the question of joint liability on joint returns and ruled that “spouses are not jointly and severally liable for a deficiency arising entirely out of the separate income of one of them.” Louis and Frida Cole filed a joint income tax return for 1929. At that time, a husband and wife could choose to include their respective incomes on a joint return and be subject to tax on the aggregate income; deductions and credits to which either were entitled would be taken from the aggregate income. After the Coles filed their joint return, the Service determined a deficiency in the amount of $27,568. The petitioner, who was the executrix of Mr. Cole’s estate, argued the husband and wife were each “liable only for the proportion of the tax attributable to his share of the aggregate income.” The Commissioner, following the Bureau’s administrative practice of imposing joint and several liability as set out in I.T. 1575, argued that it was “impossible” for the Service to prorate the liability because the liability had been calculated on their combined income. The Board of Tax Appeals had previously rejected the Commissioner’s argument that a husband and wife filing jointly became “a single taxing entity, and as such, ‘the taxpayer.’” The Commissioner, in his brief in Cole, set forth the position that “the husband and wife, must of course, accept this unusual privilege of reducing their taxes as it exists with all of its necessary concomitant conditions and results.” The Ninth Circuit did not agree with the Commissioner, finding that any administrative difficulties of apportioning the liability could be overcome; the court pointed out that, in this case, the apportioned liabilities had in fact been stipulated.
Despite the outcome in Cole, the Commissioner subsequently persisted in arguing that joint and several liability must attach to a joint filing, as a matter of administrative convenience. With the enactment of the Revenue Act of 1938 (hereinafter “1938 Act”), the Service finally got its wish: Congress added subsection 51(b), which stated “the liability with respect to the tax shall be joint and several.” After the 1938 Act, the Commissioner then began arguing in ongoing litigation involving tax years prior to the effective date of the 1938 Act that § 51 “should be deemed merely declaratory of existing law under prior revenue acts.”
In Cole, the Commissioner noted the taxpayers reduced their taxes by filing jointly, with the tax being computed on their aggregate income. At the time of Cole, joint filing produced a reduction in tax only when one spouse had deductions (including losses) or credits that could offset the income of the other. Unlike our present system, in which there are five separate filing statuses and four accompanying rate structures, the tax rates prior to 1948 were uniform for all types of returns. Thus, for many married couples, the effect of aggregating income was to push the couple into a higher tax bracket, resulting in a higher liability than if the couple had filed as two individuals. A couple would not benefit from joint filing unless one spouse had deductions or credits that could offset the income of the other spouse.
In unpacking the rationales of joint and several liability, it is important to recognize that the reduction and ensuing “unusual privilege” referred to by the Commissioner in Cole should not be confused with so-called “income splitting,” which was not a benefit of joint filing until the Revenue Act of 1948 created a separate rate structure for married couples. Somewhat ironically, income splitting was a phenomenon that was first enjoyed by married couples who filed separate returns in community property states. Under the theory that income earned by one spouse was community property of the other spouse, in 1930 the Supreme Court ruled in Poe v. Seaborn that spouses were permitted to file separately and each treat one half of the spouses’ community property income as his or her income for tax purposes. Thus, if the husband earned $10,000, and the wife did not have any income, each would file a separate return reporting $5,000 of income. Joint and several liability did not attach, because the spouses filed separately. Yet due to the nature of the progressive rate structure, the taxpayers benefitted from the lower rates applicable to the lower income amounts.
For a time following Seaborn, only taxpayers who lived in community property states enjoyed the benefits of income splitting because each individual was subject to tax on one-half of the couple’s total income. At the same time, taxpayers in common law states who filed a joint return had to aggregate their income, but were subject to tax under the same “steeply progressive” rate structure. Congress was concerned when a number of states, concerned about the unequal tax treatment of their residents, moved to adopt the community property system. To address this disparity, the Revenue Act of 1948 introduced a new rate schedule for married taxpayers filing a joint return.
So while Congress established joint and several liability for married filers in 1938, it was not until ten years later, in 1948, when the new separate rate schedules were introduced, that most taxpayers had an economic incentive to file a joint return. Significantly, the new rate schedules benefitted all married taxpayers versus their unmarried peers, because at that time the brackets were exactly twice as wide as the rate schedule for unmarried taxpayers.
As a result of this structure, a married couple was not worse off in 1948 under the married filing jointly rates than under the individual rates. One spouse’s deductions could be used to offset the income of the other (as in Cole), without the couple facing higher rates (as they had prior to the Revenue Act of 1948). The benefits of income splitting enjoyed by married couples were especially favorable when one spouse (at that time, almost certainly the husband) worked and the other spouse did not. This phenomenon became known as the “marriage bonus.”
Over time, Congress recognized that the pendulum had swung too far in favor of married couples. The Tax Reform Act of 1969 created a new, lower rate schedule for unmarried persons to correct the fact that, due to income splitting, an unmarried person’s tax was as much as forty-two percent higher than the tax paid by a married couple filing jointly reporting the same total income as the unmarried person. The Joint Committee on Taxation explained in its report accompanying the Act:
With the new rate schedule for single persons, married couples filing a joint return will pay more tax than two single persons with the same total income. This is a necessary result of changing the income-splitting relationship between single and joint returns. Moreover, it is justified on the grounds that although a married couple has greater living expenses than a single person and hence should pay less tax, the couple’s living expenses are likely to be less than those of two single persons and therefore the couple’s tax should be higher than that of two single persons.
With that rationale underlying the Tax Reform Act of 1969, the “marriage penalty” was born. However, Congress did not reconsider the fairness of joint and several liability—if the “privilege” of income splitting was now eroded for many couples, should a spouse still be held liable for the other’s tax liability? Significantly, Congress seems to envision two types of households in its thinking—that of an unmarried single person, and that of a married couple.
What about an unmarried couple living together, sharing household expenses, or any number of nontraditional household units? Demographically, the number of those households may have been statistically less significant in 1969, but today this is not the case. The rate of cohabitation in the United States is fourteen times higher in 2010 than it was in 1970, and approximately twenty-four percent of children are born to cohabiting couples. The legislative intent of certain Code provisions thus seems antiquated when viewed in light of the nation’s changed demographics.
B. Examining the Stated Rationales of Joint and Several Liability: Does the Intent Still Make Sense in Twenty-First Century America?
1. Administrative Ease for Taxpayers and Government
Dating back to the Commissioner’s briefs in Cole and continuing today, administrative ease has been an oft-cited rationale in support of both joint return filing and joint and several liability. The legislative history of the Revenue Act of 1938, picking up on the arguments made by the Commissioner in Cole and earlier cases, stated:
Unless the husband and wife are to be held jointly and severally liable for the tax upon their aggregate net income, it will be necessary for the [IRS] to require that their individual incomes and deductions shall be separately stated in the return, in order that their respective income-tax liability may be separately determined. Such a requirement would cause considerable hardship upon taxpayers with moderate incomes and would largely eliminate the advantages of the joint return.
It is true that if the married filing jointly status were eliminated, there would be a sharp increase in the number of returns that the Service must process. It is also true that would create an added burden for married taxpayers, who must prepare, or pay for the preparation of, two individual returns rather than one joint return.
However, technology has come quite a long way since 1938. The Service does not process returns by hand, nor do most taxpayers prepare a return with pencil and paper. Since 1986, the Service has encouraged (and, more recently, mandated in some cases) the electronic filing of income tax returns. These efforts are working. In June of 2011, the Service announced that it has processed more than one billion total individual tax returns through its electronic filing program. This statistic included more than 100 million individual tax returns (representing 79% of individual income tax returns filed) that were e-filed during the 2011 filing season. Contrast this with tax year 1998, in which only 19% (25 million out of 123 million) of all individual income tax returns were e-filed.
In her 2005 Annual Report to Congress, Nina Olson recommended that Congress eliminate joint and several liability and instead require married taxpayers to file a split-column tax return. She noted that Congress had cited “administrative” reasons when it imposed joint liability in 1938, and that administrative challenges had again been cited as a hurdle by the Treasury Department in its 1998 study. Her 2005 report described updates to the Service’s computer system and the rise of e-filing as two reasons that those administrative challenges were less significant in 2005 than they were even in 1998. She further suggested that, to the extent these administrative challenges remain, they “could also be partially offset by the reduction in, or elimination of, innocent spouse relief” and several other related issues that arise from joint filing.
One issue of concern is that the trend towards e-filing, however positive for the Service, creates a new difficulty in the joint and several liability context because an actual signature is not required on an electronically filed return. This facilitates the ability of one spouse to enter the tax information and file the return without the other spouse ever having the opportunity to review or sign the return. Similarly, a spouse who reviews and approves a joint return would not know if the other spouse makes subsequent changes prior to filing it. Given the structural incentives to file jointly, I believe there will be an increase in this type of “forgery” as e-filing becomes the norm. If the other spouse objects to the filing, the Service must undergo a factual determination as to whether the filing spouse forged a signature or merely acted with tacit consent based on past practice within the marriage. This is another example of how joint liability creates an additional burden on Service resources.
The rise of e-filing thus provides two important reasons in support of Congress eliminating joint filing: (1) in light of this technological advance, “administrative ease” is no longer a convincing rationale in support of joint return; and (2) it can be very difficult to ensure that both spouses reviewed and consented to a joint return that is filed electronically.
2. Joint Liability as the Price for the “Privilege” of Filing Jointly
The “privilege” of filing a joint return is frequently invoked as a justification in support of joint and several liability. Interestingly, the use of the word precedes the introduction of income splitting and the concept of the marriage bonus. It dates back at least to Cole, when the Commissioner spoke of the “‘unusual privilege’ of filing a joint return.” The legislative history accompanying the Revenue Act of 1938 echoed the Commissioner’s use of the word “privilege”:
Section 51(b) of the bill expressly provides that the spouses, who exercise the privilege of filing a joint return, are jointly and severally liable for the tax computed upon their aggregate income. It is necessary, for administrative reasons, that any doubt as to the existence of such liability should be set at rest, if the privilege of filing such joint returns is continued.
Courts emphasized this “privilege” after Congress first introduced innocent spouse relief in 1971. In its oft-cited Sonnenborn opinion released that same year, the Tax Court wrote of the privilege as a fact that must be kept in mind when weighing innocent spouse relief:
It is important that these provisions be kept in proper perspective. The filing of a joint return is a highly valuable privilege to husband and wife since the resulting tax liability is generally substantially less than the combined taxes that would be due from both spouses if they had filed separate returns. This circumstance gives particular emphasis to the statutory rule that liability with respect to tax is joint and several, regardless of the source of the income or of the fact that one spouse may be far less informed about the contents of the return than the other, for both spouses ordinarily benefit from the reduction in tax that ensues by reason of the joint return . . . . [I]t must be kept in mind that Congress still regards joint and several liability as an important adjunct to the privilege of filing joint returns, and that if there is to be any relaxation of that rule the taxpayer must comply with the carefully detailed conditions set forth in section 6013(e).
These words, written more than forty years ago, have been cited with frequency. As discussed in Part III, the innocent spouse provisions have changed considerably since 1971. But that is not all that has changed, in the Code or in society generally. If a “price” is to be paid by taxpayers for this “privilege,” we must consider to what extent and for whom joint filing remains a privilege today.
In the intervening six decades since income splitting was introduced, two major changes to the Code have occurred that erode the “unusual privilege” argument: (1) the introduction of the marriage penalty; and (2) the introduction of the earned income credit. Changing demographics exacerbate the effects of these changes to the Code. If the privilege represents the reduction in tax achieved by joint filing, then it is only a privilege to the extent that such reduction can be accomplished. The so-called “privilege” of filing jointly has no actual value to many married taxpayers versus similarly situated taxpayers who are unmarried and file as single. While the joint filing status may be a “highly valuable privilege” for some, it is certainly not the case for all taxpayers.
Income splitting is one phenomenon that results in reduction of tax, and is most pronounced in cases where one spouse earns a high income and the other does not work. However, a couple in which both spouses earn a high income is likely to face not a marriage bonus but a marriage penalty; in these cases, the spouses would owe a smaller aggregate tax liability if they were permitted to file as unmarried individuals. Arguably, there is no “privilege” for two high-earners who must file as married, yet they are held jointly and severally liable for filing jointly.
A low-income couple faces a similarly illogical conundrum: their benefit from income splitting is likely nominal as compared to a one-earner high income couple such as the Bozicks, but they must accept joint and several liability as a condition of receiving the earned income credit. At the same time, a similarly situated unmarried couple would receive a higher aggregate refund than the married couple. Joint filing cannot be said to be a “highly valuable privilege” if the resulting refund is less favorable. Part V explores this conundrum in greater detail.
- Relief is Available to Some Innocent Spouses, But at What Cost to the System?
The process affording relief to innocent spouses has become increasingly liberalized over the past forty years, allowing a requesting spouse broader grounds for relief and appeal. This is well intended; if there is to be a system for relief, it should be a comprehensive process that seeks the most equitable outcome. But as we consider the state of the relief process today, let us also flip the “highly valuable privilege” theory of joint liability on its head.
Considerable resources are devoted to the resolution of these highly factual “he-said, she-said” marital disputes each year. Are these costs really worth the benefits of joint and several liability to the government? In other words, would the Service be better off if Congress repealed § 6013(d)(3) and adopted a system of separate rather than joint liability for spouses? Given the divorce rate, among other factors, would it not be easier and more efficient to treat each taxpayer as responsible for his or her own income tax liability? An overview of innocent spouse relief will provide context to these questions.
A. Evolution of Innocent Spouse Relief: 1971-1998
For decades prior to 1971, spouses who filed joint returns were held jointly liable for any resulting deficiency or underpayment. To avoid joint liability, a spouse had to show that there was no valid joint return. Forgery was one possible defense; where a court found that a signature was forged, it concluded that there was no joint return filed. Another possible defense was duress, with taxpayers arguing that a return is not joint in the absence of both parties’ voluntary consent. Courts acknowledged that duress would be a defense to a joint liability if it was shown that the duress existed at the time the return was signed; however, courts defined duress narrowly and taxpayers were typically unsuccessful in their duress claims.
In cases where a spouse willingly signed the return, courts had no avenue to provide relief from joint liability, no matter how sympathetic the facts. For example, in Scudder v. Commissioner, which predated the enactment of the first innocent spouse provision, the petitioner signed a joint return but was unaware that her husband had embezzled money and not reported it on the return. The Service held her jointly liable for the resulting deficiency, and the Tax Court, “with considerable reluctance,” upheld the determination:
Although we have much sympathy for petitioner’s unhappy situation and are appalled at the harshness of this result in the instant case, the inflexible statute leaves no room for amelioration. It would seem that only remedial legislation can soften the impact of the rule of strict individual liability for income taxes on the many married women who are unknowingly subjected to its provisions by filing joint returns.
This plea did not go unheard. The Innocent Spouse Act of 1971 was said to have reflected the concern of Congress “about the ‘grave injustice’ imposed by joint and several liability in cases in which, for example, the culpable spouse embezzles funds, fails to report the proceeds, deserts the innocent spouse, and squanders the funds.” The legislative history to the 1971 Act recognized the inequitable outcomes that sometimes resulted from strict joint and several liability. The Senate Report on the bill noted that “some . . . judicial decisions have carried pleas for legislative relief” and quoted the above language from the Tax Court’s opinion in Scudder as an example. The report stated that “[t]his proposal seeks to correct the unfairness in the situations brought to the attention of this committee and to bring government tax collection practices into accord with basic principles of equity and fairness.”
To address these concerns, the 1971 Act added subsection (e) to § 6013, providing potential relief for a spouse who had filed jointly if: (1) there was omitted income in excess of twenty-five percent of the amount of the gross income reported on the return; (2) this omitted amount was attributable to one spouse; (3) the other spouse establishes that he or she did not know of, and had no reason to know of, such omission; and (4) it would be inequitable to hold the other spouse liable for the deficiency attributable to such omission, taking into account all facts and circumstances, including whether or not the other spouse significantly benefited directly or indirectly from the items omitted from gross income.
The notion of an “innocent” spouse has evolved considerably over time, just as the statutory provisions for relief have been liberalized. To qualify for relief under the 1971 Act, one truly had to be uninvolved in, and unaware of, his or her spouse’s financial activities and resulting understatement of income on the joint return. Moreover, relief was only available if the spouse omitted twenty-five percent or more of his or her income on the return; an otherwise innocent spouse had no relief if the omission were smaller. Since its narrowly defined introduction in 1971, the grounds for innocent spouse relief have been increasingly liberalized though legislative amendment and case law. Contrast the 1971 standard for innocent spouse relief with that now available under § 6015(f), enacted in 1998, which affords a taxpayer relief if, “taking into account all the facts and circumstances, it is inequitable to hold the individual liable for any unpaid tax or any deficiency (or any portion of either).”
Over time, Congress recognized that the early versions of innocent spouse relief were “narrowly drawn and strictly interpreted” and that many former spouses did not qualify for those protections. As part of the 1996 Taxpayer Bill of Rights 2 (TBOR2), Congress directed the Treasury Department and U.S. Government Accountability Office (GAO) to study whether the then-current innocent spouse provisions were providing “meaningful relief in all cases where such relief is appropriate.” It also called for studies of: (1) the effects of changing the liability on joint returns from joint and several liability to proportionate liability (meaning “each spouse would be liable only for the income tax attributable to the income of each spouse”); and (2) the effects of providing that the Service must be bound in its collection by the allocation of joint liability agreed upon in a divorce decree.
Prior to the passage of TBOR2, the Treasury Department requested public comment on these questions. In response, the American Bar Association (ABA) and the American Institute of Certified Public Accountants (AICPA) were among the groups that submitted proposals suggesting how joint and several liability should be replaced with proportionate liability.
In February 1998, the Treasury Department issued its report to Congress. The report concluded that the existing innocent spouse provisions “may not provide ‘meaningful relief in all cases where such relief is appropriate,’” and it carefully considered the alternatives to joint and several liability, including the ABA and AICPA proposals. However, the report ultimately rejected those proposals and recommended further modifying the innocent spouse provisions “to accommodate more cases” rather than making fundamental changes to the joint and several liability standard.
Congress undertook this task in 1998. In January and February of that year, the Senate Finance Committee held hearings in which innocent spouse reform was one of a number of items examined. The resulting compromise between the House and Senate, § 3201 of the IRS Restructuring and Reform Act of 1998, became the new § 6015 of the Code. The new section provided alternative forms of relief in three different subsections: 6015(b), (c), and (f). Section 6015(b) is a liberalized version of the “traditional” innocent spouse relief applicable to joint filers who could establish that they had no actual or constructive knowledge of the item for the understatement from which they sought relief.
Section 6015(c) allows an innocent spouse to elect an allocation of the deficiency such that his or her liability is limited to the portion of the deficiency attributable only to items allocable to the taxpayer. The electing spouse must be widowed or divorced, legally separated, or living apart (for at least twelve months) from the spouse with whom he or she filed the joint return. The subsection includes anti-abuse provisions to address fraudulent transfers of assets. The innocent spouse cannot have had actual knowledge that the item giving rise to the deficiency was reported incorrectly; however, unlike with traditional innocent spouse relief and the expanded § 6015(b), § 6015(c)(3)(C) puts the burden of proof on the Service to show that the requesting spouse had actual knowledge. Moreover, a requesting spouse who had actual knowledge can overcome this requirement if the spouse is able to establish that he or she signed the return under duress.
In addition to these two specific, carefully designed statutory provisions, Congress also added § 6015(f), which provides the Treasury Secretary authority to prescribe procedures for relief from joint and several liability if, “taking into account all the facts and circumstances, it is inequitable to hold the individual liable for any unpaid tax or any deficiency.” A taxpayer is not eligible for relief under § 6015(f) unless relief is unavailable under §§ 6015(b) and (c). Significantly, § 6015(f) is the only provision that contemplates relief for an innocent spouse in the case of an underpayment. An underpayment can result when the taxpayer files a correct return indicating a liability due, but encloses only partial payment or no payment at all. The conference report explains the intention “that equitable relief be available to a spouse that does not know, and had no reason to know, that funds intended for the payment of tax were instead taken by the other spouse for such other spouse’s benefit.”
Given its statutory authority under § 6015, the Service developed administrative guidance for how it would grant § 6015(f) relief. Over time, a combination of this and subsequent administrative guidance, Tax Court opinions, and minor amendments to § 6015 have contributed to a further liberalization of the rules favoring innocent spouses.
B. A Long Way Since Scudder – Further Liberalization of Innocent Spouse Relief: 1998-2012
The 1998 Act was intended to expand innocent spouse relief and “provide equitable relief in appropriate situations.” Over the past fourteen years, the Service has undertaken administrative efforts to further improve the process. To name just a few examples, the Service: (1) developed and revised Form 8857; (2) centralized the processing and review of the requests in one location—the Covington, Kentucky, Innocent Spouse Unit; and (3) revised letters, publications, and online tools to make married taxpayers more aware of the consequences of joint filing and the availability of innocent spouse relief. And in the past fourteen years, spouses who feel it is unfair to be held jointly liable have responded in kind. The Service receives an estimated fifty thousand requests for innocent spouse relief annually and grants fewer than half of these requests. 
Following its grant of authority in § 6015(f), the Service developed guidance explaining its standards for determining equitable relief. Revenue Procedure 2003-61 set out seven threshold eligibility requirements for equitable relief and specified circumstances in which equitable relief ordinarily will be granted. Taxpayers who meet the threshold eligibility requirements but not the specified conditions of section 4.02 might still qualify for relief; the Service can consider all facts and circumstances and grant relief if it finds it inequitable to hold the requesting spouse liable for all or part of the deficiency or underpayment. The balancing factors include: marital status; economic hardship; legal obligation of the nonrequesting spouse; whether the requesting spouse benefited significantly from the unpaid liability or item giving rise to the deficiency; the requesting spouse’s compliance with income tax laws; spousal abuse; and mental or physical health of the requesting spouse.
Interestingly, another of the balancing factors that the Service will consider in determining equitable relief is whether the requesting spouse knew or had reason to know of the underpayment or item giving rise to the understatement. This is a significant departure from the pre-1998 innocent spouse relief statute, which required a spouse to show no knowledge or reason have knowledge as one of several conditions for relief. Under the equitable relief standard, the Service (or the Tax Court, in its review) can now grant relief from liability to a spouse who had full knowledge of the wrongdoing when he or she signed the return if it feels that the other factors supporting relief warrant it.
Returning to the Bozick case as an example: both the Service and the Tax Court concluded that Ms. Bozick had reason to know that her husband would not pay the $137,453 liability shown on their joint return. The record shows Ms. Bozick was aware that he was in failing health, had credit card debt, and had a gambling problem. She acknowledged these facts in her divorce petition and also acknowledged that he did not keep up with his quarterly taxes and withholding. The Tax Court weighed all facts and circumstances in its opinion, including the relief factors listed in Revenue Procedure 2003-61, but was perhaps most persuaded that it was fair to grant her relief by the fact that “Bozick signed a joint tax return only after having been browbeaten into doing so by her husband and after having filed a separate return.”
A lot of open questions remained after the 1998 Act, even after the Service developed its administrative guidance for granting relief. For example, in the years immediately following the 1998 Act, the Service contended that the Tax Court lacked jurisdiction to review a denial of equitable relief brought under § 6015(f). The Tax Court found that it did have jurisdiction over these cases. However, in 2006, the Ninth Circuit held that the Court lacked jurisdiction in § 6015(f) underpayment cases where the liability did not arise from a deficiency. In December 2006, Congress settled this issue by amending § 6015(e) to make clear that the Tax Court had jurisdiction to review all denials of relief, including § 6015(f) underpayment cases.
The scope and standard of judicial review have been the subject of litigation in recent years, with the Tax Court holding that the scope of review could include new evidence introduced by the taxpayer at trial that was not in the administrative record and that the proper standard of review in all § 6015 relief cases is de novo. Despite Tax Court decisions to the contrary, the Service continues to argue that the scope of review of § 6015(f) cases should be limited to the administrative record and that abuse of discretion is the proper standard of review.
The cumulative impact of the 1998 Act, the 1996 amendment to § 6015(e), and the Tax Court holdings on the scope and standard of review has been a great liberalization and expansion of innocent spouse relief over a fourteen year period. This was, of course, what Congress had intended when it decided in 1998 to broaden the relief rather than eliminate joint and several liability. But to achieve “equitable” outcomes, each determination necessarily becomes a very fact-intensive inquiry into the personal lives and marriage of the taxpayers. If the nonrequesting spouse responds or intervenes, the Service and/or the Tax Court often must make a judgment in a “he-said, she-said” presentation by the spouses.
And as already noted, the liberalization of the statute does not mean that all or even most requesting spouses get the relief they seek. The Taxpayer Advocate’s 2011 Annual Report to Congress included an analysis of forty-three opinions from the Tax Court or Courts of Appeals involving relief under § 6015 for the period June 1, 2010 to May 31, 2011. Of these forty-three cases, most (thirty-four cases, or 79%) involved a reconsideration of the merits as to whether the taxpayer should be granted relief. Of these cases decided on the merits, the taxpayer was granted full relief in eleven of thirty-four cases (or 32%) and partial relief in seven cases (21%). The Service’s denial of relief was fully upheld by the court in nearly half the cases (sixteen of thirty-four, or 47%).
A significant number of the forty-three cases analyzed in 2011—fifteen cases, or 35%—involved procedural questions. This is common from year to year. The Lantz case sparked a closely-watched bout of litigation concerning procedure for innocent spouse relief cases, which eventually led to yet another liberalization of the relief process. At issue in Lantz was whether a requesting spouse must request equitable relief under § 6015(f) within two years of the Service commencing collection activity on the requesting spouse. The Code is clear that a taxpayer requesting relief under § 6015(b) or (c) must do so within this two-year limit, but it is silent as to whether there is a time limit for a taxpayer requesting relief under § 6015(f). Treasury Regulation 1.6015-5(b)(1), which was promulgated in 2002, extended this two-year limit to taxpayers seeking equitable relief under § 6015(f). In Lantz¸ the Service had rejected the taxpayer’s § 6015(f) claim as untimely because it was not within the two-year limit. The taxpayer argued that regulation section 1.6015-5(b)(1) was an invalid interpretation of the statute; the Tax Court agreed, found an abuse of discretion because the Service did not consider the merits of the taxpayer’s request, and ordered further proceedings to determine whether the taxpayer was entitled to relief. The Service appealed to the Seventh Circuit, which reversed and remanded in 2010. In cases that would have been appealed to a Court of Appeals that has not followed the Seventh Circuit, the Tax Court continued to hold that the two-year deadline imposed by the Treasury Regulation was invalid. The Service, meanwhile, continued to appeal in other circuits and win on the issue.
The two Lantz decisions attracted considerable attention among practitioners and academics. The Taxpayer Advocate highlighted this particular issue in her 2010 Annual Report to Congress, criticizing the two-year limit and calling upon Congress to amend § 6015(f) to clarify that a taxpayer may request equitable relief at any time before expiration of the period of limitations on collection. Her recommendation gained momentum in Congress, with members of the House and Senate sending letters to IRS Commissioner Schulman calling upon the Service to reconsider the Treasury Regulation in light of the intent of the 1998 Act. Senator Baucus, who was a member of the Finance Committee at the time of the 1998 Act, stated in a press release that accompanied one letter to Shulman:
We made important changes to protect innocent taxpayers seeking relief from their spouses’ liability to ensure fair and equitable treatment. Now, we are concerned this two-year limitation denies relief to the very taxpayers the law was designed to help—the innocent spouses unaware of these IRS collection activities because of intimidation or deception by their spouse. We must reevaluate these limits so all taxpayers are treated justly and have time to file for tax relief they deserve.
The letters proved effective. In July 2011, the Service announced that it would eliminate the two-year requirement in § 6015(f) cases; it will consider requests for equitable relief if the statutory period for collection remains open. Commissioner Shulman stated in the accompanying press release: “today’s change will help innocent spouses victimized in the past, present, and future.” In the year prior to this change in policy, over fifteen hundred requests were disallowed as untimely because relief was not requested within the two-year limit. The new guidance is a taxpayer friendly outcome—and good for a spouse like Ms. Lantz, whose case would have granted relief on the merits if it had been requested timely—as it will once again broaden the circumstances for relief. An accompanying increase in requests, however, will further burden a system that already receives fifty thousand such requests annually.
A few months later, in January 2012, the Service released Notice 2012-8, which introduced a proposed revenue procedure that would update and supersede Revenue Procedure 2003-61. In the press release accompanying the notice, IRS Commissioner Doug Shulman stated, “The IRS is significantly changing the way we determine innocent spouse relief. These improvements should dramatically enhance our process to make it fairer for victimized taxpayers facing difficult situations.” The notice proposes a procedure for certain streamlined case determinations, which sounds as though the Service intends to add efficiency to the process. However, it does not specify what is meant by “streamlined.” At the same time, another section of the proposed revenue procedure expands how the Service can weigh abuse and financial control in cases involving an underpayment; while this potentially broadens the grounds for relief for innocent spouses, it could also make the determination even more fact-intensive.
The innocent spouse relief process has evolved considerably since its debut in 1971. Relief has been expanded to include more taxpayers, and the taxpayers have been granted more extensive rights of appeal. From a due process standpoint, these evolutions are admirable. Congress is sympathetic to the plight of the “innocent spouse,” as was made clear in the hearings and legislative history relating to the 1998 Act and again recently by Senator Baucus. So in light of the complicated and time-consuming relief process, why has Congress chosen not to fix the root of the problem by eliminating joint and several liability? Part IV will offer some perspective, with the benefit of hindsight.
IV. Solutions – Revisiting the Treasury Department’s Recommendation to Expand Innocent Spouse Relief Instead of Eliminate Joint and Several Liability
In 1996, Congress directed the Treasury Department to conduct a study of four issues related to joint returns. The three that are relevant to this article included: (1) the effects of changing from a joint and several liability standard to a proportionate liability standard; (2) the effects of allowing the Service to be bound in collections efforts by the allocation of joint liability in a divorce decree; (3) whether the innocent spouse provisions in effect at the time “provided meaningful relief in all cases where the relief was appropriate.”
The Treasury Department report, which was released in February of 1998, acknowledged that the existing joint and several liability and innocent spouse rules were “imperfect in certain respects.” It considered the pros and cons of a number of various proposals to modify or eliminate joint and several liability, and concluded that each of the proposals had both advantages and defects. The Treasury report opted instead to recommend legislative changes that would “preserve the advantages of the current system and yet afford innocent spouse relief in more situations.” Specifically, it recommended legislation that would broaden relief in the following ways:
Automatically suspend collection efforts against one spouse when the other is contesting a proposed joint assessment in Tax Court; Make innocent spouse relief easier to obtain by changing statutory standards to help additional taxpayers, including those with smaller tax bills who are presently ineligible for relief in many cases; Give more taxpayers who are denied innocent spouse relief by the IRS an opportunity to appeal the decision to Tax Court, and automatically suspending collection while the Tax Court considered the appeal.
Before reaching that conclusion and making those recommendations for legislative change, the Treasury report considered several alternatives. The report considered the specific advantages and disadvantages of each proposal, including the proposal that I advocate for in Part V of the article: to mandate separate returns for all married individuals, thereby eliminating both joint and several liability and joint returns.
A. A Systemic Change: Mandatory Separate Returns for all Fliers
The report considered the advantages and disadvantages of adopting a mandatory separate return system from the perspective of both the taxpayer and the Service. Mandated separate return filing is my favored solution because I believe it is unfair to force married filers to choose between joint liability and unfavorable tax treatment. The report commented in a footnote that mandated separate filing would “arguably be inconsistent with horizontal equity, one of the economic goals of taxation, because similar families would be taxed differently based on the division of income between spouses.” As I argue below, this argument is antiquated in that it presumes that a family unit consists of a married couple.
Another primary concern the report raised for taxpayers is that mandated separate filing would increase the administrative burden for couples, who would have to file two tax returns annually instead of one joint return. The report noted that this would be especially burdensome for couples with jointly held income-producing assets or jointly held assets, such as a primary residence, that create a deduction. It noted that because a comparatively small number of taxpayers choose the “married filing separately” status, questions arising in those cases were addressed on a case-by-case basis, but that if mandated separate filing were adopted, then “[c]omprehensive rules would need to be implemented, most likely through legislation, to instruct taxpayers and the Service on how to allocate any number of items between the spouses.” Perhaps because the numbers are insignificant relative to joint filers, the report does not address the situation in which unmarried people own income-producing property jointly, yet file an individual return.
From the perspective of IRS administration, the report noted that mandatory separate filing would be burdensome insofar as the Service would have to process forty-nine million additional tax returns annually. However, as I discussed in Part II.B, the IRS e-file program has helped to streamline return processing, and the number of taxpayers participating in the e-file program has nearly quadrupled since 1998.
B. Less Comprehensive Proposals: Proportionate Liability, Allocated Liability, and Adoption of the Divorce Decree
Another line of proposals in the Treasury report contemplated preserving joint filing while allowing married couples to separate their joint liability. For example, the ABA Tax Section Committee on Domestic Relations proposed what became known as “back end proportionate liability,” whereby a married couple could allocate and separate a tax liability or assessment in two situations: (1) upon election by one spouse in the case of an underpayment; and (2) “upon the assertion of a deficiency of tax” during an exam. Thus, married couples would continue to enjoy the benefits of income splitting but in many cases could avoid joint liability attributable to the tax items of the other spouse. AICPA introduced a different proposal, which became known as the “Allocated Liability Standard,” which would permit married taxpayers to file jointly and provide, at their complete discretion, their own agreed-upon allocation of the aggregate liability at filing. The report rejected both the ABA and AICPA proposals on the grounds that neither would fully eliminate the need for equitable or innocent spouse relief.
A third proposal would require the Service to be bound by the terms of a divorce decree regarding the responsibility for tax liabilities on prior joint returns. Congress had considered such a proposal previously but was troubled by the fact that the Service is not a party to the divorce proceeding.
C. Judging With the Benefit of Hindsight – Did Treasury Make the Right Recommendation in 1998? The Unending Inefficiencies of the Innocent Spouse Process
Ultimately, in 1998 Congress followed the report’s recommendations and retained joint and several liability while expanding the grounds for innocent spouse relief. In hindsight, was this the right decision?
The report raised legitimate concerns about the increased administrative costs to both the taxpayer and the Service if a mandatory separate filing service were adopted. However, since the time of the report, e-filing has changed the administrative landscape considerably. To the extent that return processing is increasingly automated, the number of returns filed is less relevant than it once was.
Meanwhile, the expansion of innocent spouse reform has broadened the grounds for relief and created new administrative burdens. The expansion has resulted in a corresponding increase in the number of requests for innocent spouse relief, with each request requiring a time-consuming and personalized determination. The Service and Tax Court are now expending significant resources to investigate claims of spousal abuse, economic hardship, and other highly factual non-tax questions such as an individual’s level of knowledge (i.e., which spouse knew what facts about items of income).
Unsurprisingly, given the task at hand, the Service has not been getting the results right. In 2007, the Treasury Inspector General for Tax Administration (TIGTA) performed an audit to examine certain procedures that had been established to protect the rights of spouses requesting Innocent Spouse relief. Specifically, the procedures were designed to ensure that collection enforcement was suspended against the requesting spouse (but not against the nonrequesting spouse) while the relief request was pending; for example, while the request is under consideration, the requesting spouse’s refund should not be offset to satisfy the joint liability that is at issue. However, TIGTA found that in twenty-seven percent of the accounts it reviewed, the Service had not acted timely to ensure that the proper protections were in place.
Richard C.E. Beck, who made a compelling argument for the repeal of joint and several liability in a 1990 article, re-examined the issue following the Restructuring and Reform Act of 1998. In his re-examination, Beck describes the high administrative costs of expanding innocent spouse relief in 1998 (as compared to abolishing joint and several liability). Beck cites IRS statistics describing how the Service created a central processing center for Innocent Spouse claims and assigned 953 full-time employees to work exclusively on those claims; despite this redirection of resources, it took the Service an average of 192 days to process claims for relief that were granted at the administrative level and it took an astounding 807 days to process claims that were appealed within the Service. Within the context of the significant resources the Service redirected toward innocent spouse relief, Beck examined the innocent spouse cases that were litigated in 2005 and found the following:
Summing up, it seems fair to say that there are now more cases than before the 1998 reforms, and that the law and the decisions are more complex, but that there is no more certainty, consistency or rationality in the case law than before. And because the ratio of wins and losses appears very similar in samples taken nearly twenty years apart, one may perhaps conclude also that the quality and consistency of the administrative dispositions are also no better now than before.
Beck argues that joint and several liability should be repealed because it is a system that lacks any justification and results in an intolerable “tax persecution of women.” He points out that “[t]he United States is virtually the only country in the developed world which insists upon [joint and several liability], even among countries which permit income-splitting to married persons filing jointly.”
I agree with Beck that the 1998 expansion of innocent spouse relief was a failure. I believe that the Service should focus its resources on the proper administration of the tax law. Congress should not expect the Service to investigate or judge intimate questions of family relations. Similarly, these cases are a burden on the Tax Court’s docket, and the court is no better suited to determine “innocence” than the Service is. The extensive litigation and advocacy regarding the two-year rule in Lantz and its companion cases is only the latest example of the tremendous resources that have been poured into innocent spouse relief. Beck argues that the innocent spouse rules will never be satisfactory, because “innocence is an irrational basis for relief and can never be made rigorous or consistent.”
In response to such critiques by Beck and other scholars, Stephanie Hunter McMahon conducted an empirical study of innocent spouse decisions to determine whether courts are implementing Congress’s legislative intent. She aggregated the results from 444 cases and concluded that “courts apply their own interpretation of congressional intent, only loosely confined by the terms provided by the executive agency.” McMahon elaborates on the courts’ interpretation of the factors set out by the Treasury Department in Revenue Procedure 2003-61 and describes some of the trends she found in her study. She calls for clearer guidelines to be established by Congress and definitions by the Treasury Department, finding it problematic that “in their opinions judges are neither crafting precise definitions of many of . . . [the terms described in the Revenue Procedure] nor defining the relative importance of each.”
McMahon does not agree with Beck’s statement that innocent spouse relief has “degenerate[d] into a global subjective test of whether the spouse seeking relief can move the judge to sympathy,” but she does conclude from her study that litigating innocent spouse relief “can be complicated and costly for both taxpayers and the government” and that “[as] with all equity claims, the factors considered by the courts may be inconsistently applied.”
In my view, the Treasury Department’s solution of “afford[ing] innocent spouse relief in more situations” has fallen short in providing relief, while simultaneously burdening the system with time-consuming claims. It is time for Congress to return to the drawing board and reconsider the message that was delivered during the 1998 testimony it heard from those who suffered because of joint liability.
Instead of expecting the Code to resolve complex family dynamics, Congress should reassess and take a bold step in the opposite direction: it should stop differentiating taxpayers according to family dynamics. Perhaps seventy-five years ago it made sense for Congress to divide taxpayers into two distinct groups based on marital status. But it is time for Congress to disentangle the Code from marital status, and this can be accomplished in harmony with a broader movement towards simplification of the Code. The solution is to ignore marital status and, in doing so, to reject the long-accepted premise that married couples behave as one economic unit.
V. Neither the Code nor America Looks Like It Did in 1938 (or Even 1975): Why Congress Should Again Revisit the Call for Mandatory Separate Filing
As described in Part IV, there are many ways to address the problems that arise from joint and several liability. Many of the solutions would not require large scale modifications to the Code. One example, contemplated by the Treasury Department Report, is that Congress could require the Service to respect the allocation of tax liability for past years agreed upon in the divorce decree. The Service could create a one-page form on which the taxpayers could allocate the liability and sign as part of divorce proceedings; once submitted to the Service, the taxpayers’ accounts could be adjusted accordingly.
This solution would be simple to enact and relatively easy to administer; it is desirable insofar as it would eliminate the need for the innocent spouse process. However, it would do so without a thoughtful reconsideration of the rationales of joint and several liability, the policy underlying marital rates, or the various grounds for penalizing taxpayers who file separately. As I will describe in this part, tax policy has not kept pace with demographic changes in the United States and there are important reasons to reconsider the Code’s underlying assumptions about households.
Moreover, the time is ripe for ambitious statutory changes. Faced with a growing deficit and following the historic U.S. credit downgrade by Standard & Poor’s, there is talk among policymakers of fundamentally revisiting the Code. Since 2010, commentators, politicians, and government officials have increasingly turned their collective attention to the national debt. Several tax policy groups, including the bipartisan National Commission on Fiscal Responsibility and Reform established by President Obama, have released tax reform recommendations. They are united in calling for major changes to the Code with a move towards simplification—closing loopholes by eliminating certain deductions in order to broaden the base and in turn reduce individual rates. In that spirit, there may be a window of opportunity for fundamental, systemic changes to the Code. The proposal to move to mandatory separate returns for all filers, a striking departure from the longstanding U.S. tax system that would eliminate the distinction between married and single filers, would fit within such a call for simplification and could be enacted as part of a broader overhaul of the Code.
It is a new century; simply put, the United States does not look like it did in 1938 when joint and several liability was adopted. Importantly, neither does the Code. There are many good policy reasons for Congress to revisit the Code’s overly simplistic assumptions about households. After a brief examination of the current demographics, I will focus on the unfair conundrum faced by low-income families headed by married couples as one way in which the married filing jointly status is particularly antiquated and counter-productive in today’s America.
A. A Closer Look at the Changing Demographics
In 1993, Marjorie Kornhauser wrote about the demographic trends at that time, which were already reflecting “the decline of the traditional nuclear family and an increase in the number of divorces, single-parent families, nonmarried cohabitation, and two-earner families.” She noted that nonmarital households increased nearly four hundred percent from 1970 to March 1991,  leading her to re-examine the concept of “family.” At the same time, Kornhauser examined the premise that married couples act as one economic unit by sharing or pooling all income. Her conclusion, based upon her own empirical study and the work of several others, was that not all married couples pool their assets and the pooling of assets is not confined to married couples. She further concluded that even among those couples who profess to pool assets, “in reality the nonearner spouse often does not have equal access to assets; instead the earner controls the money.” She concludes, as I do, that the joint return should be abolished.
In the nearly twenty years since Kornhauser’s article, the divorce rate has held steady but the percentage of unmarried cohabitating couples has risen. A recent Pew Research Center analysis of census data revealed that only 51% of adults in the United States were married in 2010. This number represents a record low and a 5% decline in new marriages from the prior year. In contrast, 72% of all adults were married in the year 1960. While the rate of marriage has dropped, the census data shows that other types of households, including cohabiting couples, single-person households, and single-parent households, have grown in recent decades. The percentage of divorced or separated adults was 14% in 2010; this number has remained somewhat steady in the past twenty years, though the number represents a stark contrast to the 5% who were divorced in 1960. Meanwhile, public opinion about marriage mirrors the decline: 39% of Americans surveyed in 2010 answered affirmatively when asked the question, “Is marriage becoming obsolete?” A greater percentage of young adults (44% of those ages 18 to 29) than older adults (32% of those over 65) agreed that marriage is becoming obsolete. A greater percentage of adults with a high school education or less (45%) agreed that marriage is becoming obsolete as compared to those adults with a college degree (27% of whom agreed with the statement).
The Pew statistics also reveal a greater divergence over time in marital rates among those with a high school education or less as compared to those with college degrees. In 1960, 72% of adults with a high school education or less were married; in 2010, only 47% of this same demographic was married. While the rate of marriage has also declined among those with a college degree, the decline is less dramatic: from 76% in 1960 to 64% in 2010.
Why are these changing demographics of relevance? One reason, which I examine in sub-part V.B, is that married low-income taxpayers are disproportionately impacted by their filing status options. A low-income married couple receives little or no benefit from filing jointly as compared to a similarly situated unmarried couple. At the same time, they must accept joint and several liability when they choose this status. If, however, the married couple files separately, there is a punitive outcome: they lose their eligibility to claim the earned income credit.
While this conclusion focuses on the example of low-income taxpayers, there are numerous other ways in which changing demographics are rendering the policy assumptions of the Code incoherent, and other scholars have argued that the joint return should be abolished for different policy reasons. For example, Lily Kahng cites changing demographics and the relative increase in the number of single people in her argument that abolishing the joint return is a matter of fairness. Anthony Infanti has also proposed mandatory individual filing, citing changing demographics in his argument for a more inclusive notion of what constitutes a family.
B. Low-Income Households Are Disproportionately Affected by the Filing Status Options for Married Persons: Why the Code Should Tax the Type of Household, Not the Relationship
Part III examined the historical rationales for joint and several liability. As the Bozick case illustrates, a one-earner married couple with a high income enjoys a tremendous benefit from the income splitting a Revisiting the Treasury Department’s Recommendation to Expand llowed by the married filing jointly rate structure. But for low-income married couples, the benefits of income splitting are not as measurable. Instead it is the eligibility for the earned income credit that is at stake by filing jointly. Even if one agrees that joint and several liability is an acceptable “price” to pay for the benefit of income splitting, one cannot ignore the fact that many low-income taxpayers do not enjoy this benefit. Yet they too are held jointly liable, and as noted previously, low-income taxpayers are the most likely to request innocent spouse relief.
As an illustration of this problem, consider the following four scenarios, each involving the same facts: a married couple has two minor children; they live together all year, which means they must file either a joint return or married filing separately. The Code does not permit them to file as single. Assume that in tax year 2010, the husband earned $25,000. The wife, who worked only part-time, earned $8,000. Both taxpayers are wage earners. The following alternative scenarios illustrate how widely the couple’s tax situation varies depending on filing status.
Scenario 1: If the couple filed “married filing jointly,” the taxpayers would be due a refund of $3,897.
Scenario 2: If the couple filed “married filing separately,” with the husband claiming the two children as dependents, the result is much different: the husband would be entitled to a refund of $1,162, while the wife would be due no refund.
As a household unit, the couple has a far less favorable tax outcome with “married filing separately” status. Their aggregate refund is $1,162, which is $2,735 less than they received as joint filers. There are two reasons for this. The primary reason is that taxpayers filing married filing separately are ineligible for the earned income credit, regardless of income level. As joint filers, they received a $2,600 earned income credit on their return. Separately, they receive nothing.
A second reason, though far less significant as a dollar amount, is that they lost the benefit of income splitting. Even though both taxpayers are in the ten-percent bracket for “married filing separately,” the wife “wastes” part of her personal exemption because she only earned $8,000. When both spouses aggregated their incomes on a joint return, the husband benefited from the portion of her personal exemption that she didn’t need. But this example illustrates how insignificant the benefit of income splitting is to a low-income couple: in this case, the tax savings attributable to income splitting was only $135—far less than the amount of the benefit that the earned income credit provided.
Put differently, in Scenario 2, the couple suffered the loss of two benefits of joint filing—income splitting and the earned income credit—with the loss of the latter costing the couple nearly twenty times more in tax savings than the loss of the former. As discussed below, the earned income credit did not exist when Congress considered the “benefits” of joint filing status and adopted joint and several liability in 1938; the lawmakers were thinking only of the ability for couples to offset one spouse’s losses against the other’s income. Income splitting became a second, justifiable rationale in 1948 when the new rate schedule for married taxpayers was adopted, creating a marriage bonus. But any marriage bonus today because of income splitting is modest for low-income taxpayers, and is dwarfed in significance by the earned income credit.
Scenario 3: Suppose there is a couple that is similarly situated to the couple in Scenarios 1 and 2, but has decided not to marry. They live together as committed domestic partners and raise their two minor children together. Their incomes are exactly the same as the other couple. First assume that each individual files as “single,” as they are entitled to do, and that the male partner (the higher earner) claimed both children: the male partner would have received a refund of $4,392 and the female partner would have received a refund of $416. As a household, the aggregate refund is $4,808, which is $911 more than the refund that the married couple filing jointly received in Scenario 1.
Why? As a single filer, each taxpayer is eligible for earned income credit based upon their income and without regard to the other taxpayer’s income. Thus, their total earned income credit as two single people is $3,646; as a married couple, it was $2,600. The male partner owes slightly more in preliminary tax liability, because he lost the $135 benefit of income splitting; however, the $1,046 increase in their household earned income credit well outweighs the household’s $135 tax increase.
Why do two unmarried taxpayers sharing a household receive a significantly higher earned income credit than a married couple, in a scenario where the household income is exactly the same? Because the earned income credit is determined by income level and filing status, not by household composition. Thus, in Scenario 3, we see a disincentive for low-income taxpayers with children to marry. In reality, Scenario 4 illustrates how the unmarried couple will likely attain an even larger household tax refund than that shown in Scenario 3.
Scenario 4: If the higher earner of the couple in Scenario 3, in this case the male partner, qualifies for “head of household” filing status, the father’s refund will be $4,662, while the female partner’s refund remains $416. In the aggregate, the couple receives a refund of $5,078, which is $1,181 more than a similarly situated married couple would receive.
The wildly disparate results shown in Scenarios 1 to 4 show that, under these facts, the hypothetical couple is better off from an income tax perspective if they are unmarried rather than married. If unmarried, they will avoid the pitfalls of joint and several liability. But if married, they cannot do so without being further penalized if they elect separate filing status. It does not make sense to incentivize a married couple to file jointly (thus introducing a potential risk to the “innocent spouse”) while allowing a similarly situated unmarried couple to avoid joint liability yet receive higher refunds.
This is one of many reasons for which I conclude that Congress must move beyond joint and several liability and joint filing. Both the country and the tax code have changed considerably since 1938, and it is time to revisit the suitability of this structure.
1. Isolating the Impact of the Earned Income Credit in the Low-Income Context
Scenarios 1 to 4 highlight some disturbing inconsistencies with the intended tax policy, all stemming from the complexity and the value of the refundable earned income credit in its current incarnation. First, why should taxpayers be ineligible for the earned income credit merely because they file separately, if they would meet the income limits for eligibility as a couple? If the couple wishes to preserve separate liability, so as to avoid the types of legal difficulties that arise under joint and several liability, then they must forsake their earned income credit eligibility. Second, why should an unmarried couple living together receive a higher total earned income credit than a similarly situated married couple? A couple sharing household expenses will face the same household costs irrespective of marital status, but the unmarried couple benefits more in this case. This scenario demonstrates a financial disincentive for the couple to marry. One reason for these inconsistencies in tax policy is that the earned income credit was introduced decades after joint and several liability, and the original purpose and scope of the earned income credit were quite different than what the program accomplishes today.
The earned income credit was first introduced in the Tax Reduction Act of 1975, which was a temporary stimulus measure enacted to address the weak economy and high unemployment at the time. The earned income credit itself was to be temporary: as originally enacted, it was effective only for tax year 1975. The legislative history describes the credit as appearing to have the same two objectives as the work bonus plan that had been proposed several times and not passed: (1) “as a way of decreasing work disincentives” for persons on welfare; and (2) as a way of addressing the regressive nature of social security taxes. The maximum earned income credit available in 1975 was $400, and it was available only to low-income taxpayers (both unmarried and married) with a dependent child in their household. The credit began to phase out if the taxpayer’s adjusted gross income exceeded $4,000, and a taxpayer with an adjusted gross income of $8,000 or more would receive no credit at all. As is still the case today, the credit as originally introduced was available to a married couple only if a joint return was filed; a couple filing separately could not receive the credit.
The earned income credit, of course, was extended beyond the 1975 tax year and has been expanded repeatedly over the years. Today it holds an incredibly significant benefit for low-income taxpayers. Unlike in 1975, it is currently available even to certain low-income taxpayers who do not have children. The credit is calculated based on income and phases out at different maximum income levels that vary depending on whether a taxpayer is unmarried or married filing jointly; the phase-out levels increase according to whether the taxpayer has no children, one child, two children, or three or more children. The amount of the maximum credit increases in the same manner. For example: in 2010, a taxpayer with no qualifying children is eligible to receive a maximum credit of $457, and the credit phases out to zero if the taxpayer’s adjusted gross income is $13,460 or more (or $18,470 if married filing jointly). A taxpayer with one qualifying child can receive up to $3,050, but will receive nothing if the taxpayer’s adjusted gross income is $35,535 or more (or $40,545 if married filing jointly). A taxpayer with two qualifying children can receive up to $5,036, but the credit phases out to zero at an adjusted gross income of $40,363 (or $45,373 if married filing jointly). Finally, a taxpayer with three or more qualifying children might receive up to $5,666, with the credit phasing out to zero at an adjusted gross income of $43,352 (or $48,362 if married filing jointly).
There is evidence that the earned income credit not only lifts families out of poverty but benefits children in more meaningful ways. For example, a recent study by Gordon Dahl and Lance Lochner estimated the impact of the increase in family income due to expansions in the earned income credit program. Dahl and Lochner concluded that their data indicated “modest but encouraging effects of family income on children’s scholastic achievement.” If one accepts this conclusion as an influencing consideration of tax policy, then it is even more nonsensical to link eligibility for the earned income credit to joint and several liability.
2. Isolating the Impact of the Marriage Bonus in the Low-Income Context
Ignoring the impact of the earned income credit, the marriage bonus is far less pronounced for low-income taxpayers, especially those who can claim exemptions for one or more dependent children, because their rates are relatively low compared to those of high-income taxpayers. Yet there are scenarios in which the marriage bonus exists for low-income couples, so it should also be examined as a “privilege” of filing jointly in this context. If Congress moved to individual filing, low-income taxpayers would stand to lose whatever marriage bonus attributable to income splitting they now enjoy.
The most noticeable marriage bonus would result in a one-earner household with no children. Staying within the context of low-income earners, imagine that one spouse earned $40,000 and the second spouse earned no income: the marriage bonus would be $1,822, all of which is attributable to income splitting and the extra personal exemption (a childless couple will receive no earned income credit at this income level). This amount is a significant sum for a couple with a household income of only $40,000. However, it is questionable as a matter of tax policy whether the Code should incentivize a one-earner household when there are no children. This scenario lacks the obvious compelling factor, namely child rearing, to justify the second earner’s lack of participation in the labor market. Meanwhile, the non-earner is left financially vulnerable in the event of divorce or the earner’s death. If Congress moved to mandatory separate filing, this couple would lose the $1,822 benefit of income splitting, but the non-earner would not bear the risk of an unforeseen individual tax liability. Additionally, the Code would no longer provide a disincentive (real or perceived) for the non-earner to enter the labor force by virtue of marital status.
The marriage bonus is less pronounced if the same couple has two children, such that one spouse earns $40,000 and the second spouse earns no income because he or she has elected to stay at home with their small children. In that case, the amount of the marriage bonus attributable to income splitting and the second spouse’s personal exemption is $1,101 as compared to an unmarried person filing as head of household. Again, while a high-income one-earner couple would receive a far greater marriage bonus, $1,101 is not an insignificant sum for a low-income couple. Ironically, if the second spouse in this example were to work part-time and earn as little as $6,000, the couple would completely phase out of earned income credit eligibility. Thus, there are two separate tax disincentives for the second spouse to stay out of the work force in this hypothetical.
In any event, the calculus regarding a marriage bonus as a disincentive to work is far more complex when small children are present, because the couple will incur day care costs if the second spouse enters the work force; the second earner will have to earn a certain minimum amount to offset those costs, while simultaneously reducing the marriage bonus (and the earned income credit) with his or her earnings. Thus, it is often not rational for low-income families with children to follow the two-earner model. However, there are other ways to address the costs of raising children, whether those costs are borne as day care expenses or the opportunity cost of the second earner staying home.
C. A Brief Consideration of How to Restructure the Earned Income Credit in an Individual Filing System
If Congress mandated individual filing, it would have to decide how to structure the credits currently available to households headed by married couples. One specific question is whether Congress would aggregate the spouse’s separate income for purposes of determining the earned income credit. As noted above, it does not now do so for cohabiting couples. Should married couples be treated differently? If one does not accept the premise of marriage as an economic unit, then arguably the Code should not aggregate the spouses’ incomes. Zelenak examined the pros and cons of this question in his 1994 article, noting that:
If a wife of a high income husband earns a few thousand dollars . . . [t]he premises of separate returns—marriage neutrality and the independent economic identities of spouses—indicate that she should be eligible for the credit. But the credit is intended to help the poor, and she is not poor.
It is true that she is not poor, but Congress should not ignore the fact that many marriages end in divorce, and she may be better positioned to support herself after a divorce if she maintains a presence in the work force throughout her marriage. Furthermore, if her high-earning spouse claimed the children as dependents on his return, as he is likely to do, then her separate entitlement to an earned income credit would be nominal.
Zelenak suggested that, were Congress to adopt mandatory individual filing, a better alternative to retaining the current earned income credit system would be to redesign the credit. As Zelenak notes, George Yin and Jonathan Barry Forman proposed a family-based credit that could work well in a separate return system. Yin and Forman contemplated a two-pronged program: an exemption from Social Security taxes for the working poor, and a refundable credit for low-income taxpayers based on the number of children in the home.
If Congress desires a substantial subsidy for low-income households with children as a matter of tax policy, as the current Code supports through refundable credits, it seems reasonable to believe that such a credit could be completely disaggregated from filing status. A married couple should not benefit less from such a subsidy than a cohabiting couple, and a wife should not bear the risk of being held responsible for her husband’s tax shortcomings in a program designed to benefit her children.
When the Senate Finance Committee held its Restructuring Hearings in 1998, Senator William Roth opened the discussion on innocent spouse tax rules by highlighting the plight of the low-income ex-wife who was struggling to make it as a single parent:
Financially insecure, many times struggling as a single parent to raise children, working for an income that is a fraction of what her ex-spouse earns, now she has to confront the often unrelenting Internal Revenue Service. In an effort to acquire revenues owed, the agency will pursue these women with a vengeance. It will garnish wages, place liens against homes, and often jeopardize future relationships because a new person in her life might well be held accountable for her former spouse’s tax problems.
Despite the extensive legislative reform that followed these hearings in 1998 and years of further statutory, judicial, and administrative liberalization of the innocent spouse rules, joint liability continues to adversely impact low-income women. Rather than continuing to expand the grounds and factors for relief, subjecting these women to an invasive relief process that takes many months if not years to resolve, Congress should revisit its antiquated assumptions about joint filing and move to mandatory individual filing.
While this will create ripple effects throughout the Code, it is also an opportunity to move towards simplification and to update the Code in light of the changing demographic realities of the United States.
Bearing in mind the continued decline of marriage rates (which is most pronounced among the least educated) and the potential impact of family income on a child, it is harmful tax policy to condition a married couple’s eligibility for the earned income credit on the acceptance of joint and several liability while asking a divorcée to turn to an inefficient relief process if the marriage ends. The better answer is to stop penalizing the innocent spouse, embrace demographic reality, and simplify the filing status structure by mandating individual filing.
Acknowledgments: I would like to extend my thanks to the Frances Lewis Law Center for research support, as well as the following people who provided helpful feedback at varying stages of the project: Robin LeBlanc,Keith Fogg, and Caroline Chen. I also thank Emerald Smith, William Harrison, and Darby Gooding for their research assistance.
 I.R.C. § 6013(d)(3) (2012).
Robert W. Wood, The Marriage Trap, Forbes, June 27, 2011, at 110, available at http://www.forbes.com/forbes/2011/0627/money-guide-11-tax-joint-return-irs-wedding-marriage-trap.html. See also Tax Policy Center, Return Details by Marital Status, 2005–2009, available at http:/www.taxpolicycenter.org/taxfacts/Content/PDF/returns_marital.pdf. The figures were provided in July 2011 by the IRS Statistics of Income Division, which reported that of 140,494,127 returns filed for tax year 2009, 53,570,158 were returns of married persons filing jointly, while 2,539,588 were returns of married persons filing separately.
 It is theoretically possible for a married taxpayer to face a lower tax liability by filing separately, but I have yet to encounter this in practice. For example, if one spouse has significant medical expenses that exceed 7.5% of his or her own adjusted gross income (AGI) but not the couple’s total AGI, it could benefit the taxpayer to file separately. However, it is possible that the taxpayer will find him- or herself subject to the alternative minimum tax, negating the possible tax advantage.
 This is true for taxpayers at both low and high income levels, though for different reasons. Low-income taxpayers cannot receive the valuable earned income credit if they choose to file separately, even if they would be eligible for the credit if they filed jointly. Married middle- and high-income taxpayers (those individuals with a taxable income of at least $68,650 in 2010) who file separately are subject to a more progressive income scale than unmarried taxpayers, in that their tax brackets are narrower between rate increases. For example, a married individual filing separately is subject to income tax at the 33% rate starting at $104,625. Contrast this to an unmarried individual, who is not subject to the 33% rate until he or she has taxable income of $171,850. At all income levels, married taxpayers lose the benefit of income splitting if they choose to file separately. Part IV will examine the disparate treatment of similarly situated individuals in greater detail with a particular focus on those individuals who are eligible to receive the earned income credit.
 I.R.C. § 6402 (2012).
 To illustrate how sizable a single parent’s refund can be, imagine a divorced mother of three young children who earns $25,000 as a wage earner. In tax year 2010, assume that the mother files as head of household, claims the children as dependents, and is eligible for both the earned income credit and the child tax credit. (These figures disregard the Making Work Pay credit, which would have been available to her in 2010). She would be due a $6,659 refund even if she had no income tax withheld from her pay in 2010. If the same taxpayer earned only $20,000, she would be due a $7,463 refund.
 A divorce agreement has no effect on the Service’s collection of a joint and several liability. The government defends this practice on the grounds that it is not a party to the divorce and that it would be impractical for the Service to become involved in divorce settlements. See, e.g., U.S. Gov’t Accountability Office, GAO/T-GGD-98-72, Innocent Spouse: Alternatives for Improving Innocent Spouse Relief 10 (1998) (“[P]roviding a legal forum where IRS and the parties to a divorce could resolve issues relating to both tax matters and divorce proceedings would require a fundamental and extensive change in either federal tax law or state domestic relations law”). Notwithstanding the Service’s disregard for the couple’s agreement, one spouse may have a legal right of contribution under the divorce decree.
 See Nat’l Taxpayer Advocate, Joint and Several Liability, FY 2001 Ann. Rep. to Cong. 139 [hereinafter TAS Report 2001] (showing that when married individuals file a joint return, the primary spouse usually earns significantly more than the secondary spouse). The table in the report shows that this is true at all income levels, but is particularly pronounced at the high end and the low end: in the “under $30,000” income range, 91% of gross income was attributable to the primary spouse; in the “over $100,000” income range, 79% of income was attributable to the primary spouse. Id. The numbers in the table are from tax year 1999. Id.
 All facts are taken from the Tax Court opinion Bozick v. Comm’r, 99 T.C.M. (CCH) 1242 (2010). This case does not hold special significance and I was not involved with this case in any capacity. I chose this case because its facts provide a good illustration of the valuable marriage bonus, the ensuing risks of filing jointly to benefit from that bonus, and the lengthy process for seeking relief from joint liability.
 A marriage bonus results when a married couple enjoys tax savings by virtue of the joint filing rates. The bonus is most pronounced when one spouse earns a high income and the other has little to no income. The opposite phenomenon is called the “marriage penalty,” which occurs when two spouses earn significant and roughly equal incomes; they pay more income tax than they would in the aggregate if they were unmarried. These concepts are discussed further in Part I.A. See also infra note 54.
 Congress first enacted innocent spouse relief in 1971 and has expanded the provisions and the scope of relief repeatedly over time. Currently there are three types of relief provided in I.R.C. § 6015. Part II provides a detailed history of innocent spouse relief.
 As I discuss elsewhere, the process for obtaining relief is typically lengthy and often involves an administrative or judicial appeal. In this case, I should note that the IRS denial came quickly and the three year period was attributable in part to Ms. Bozick having been granted a continuance for her trial date in order to obtain counsel.
 By “low-income taxpayers” I refer to those taxpayers who are income-eligible to receive the earned income credit. In 2010, a married couple with income up to $48,362 could receive the earned income credit, so long as they filed jointly and had at least three qualifying children. The income limits are lower for couples with fewer or no children.
 Section 7703(b) provides an exception for certain married individuals living apart from a spouse during the last six months of the taxable year and maintaining a household with a child; an individual meeting the requirements of § 7703(b) is not considered married and thus remains eligible for the earned income credit because he or she is permitted to file as “head of household” rather than “married filing separately.”
 See infra Part IV.B.
 See TAS Report 2001, supra note 8, at 134 (“Given the many steps and legal requirements in the process, even the best case scenario requires 304 days to process a claim for relief under IRC § 6015.”).
 See id. at 128 (stating that the IRS rejects 49% of claims for not meeting requirements for consideration and only 47.7% of the remaining claims obtain any relief). See also Nat’l Taxpayer Advocate, Another Marriage Penalty—Taxing the Wrong Spouse, FY 2005 Ann. Rep to Cong. 423 (2005) [hereinafter TAS Report 2005] (citing a 2005 IRS report that found innocent spouse relief was granted at the administrative level “on only about 30 percent of all claims”).
 References to a statute section are to the Internal Revenue Code of 1986, as amended, unless otherwise indicated.
 In many recent years, the Service has reported receiving more than fifty thousand requests for innocent spouse relief. See, e.g., Nicola M. White, Lawmakers Urge IRS to Review Innocent Spouse Relief Regulations, 131 Tax Notes Today 358 (2011); Laura Saunders, A New Push to Protect Spouses, Wall St. J., May 28, 2011, at 9-B.9. A significant number of these requests are immediately rejected for not meeting threshold criteria. Of the remaining requests, the relief rate is still low. See, e.g., Carla Fried, For ‘Innocent Spouses,’ a Helpful Shift in I.R.S. Policy, N.Y. Times, Feb. 12, 2012, at 11-BU.11 (stating that “22 percent of the 32,000 cases filed each year that qualify for review are granted full relief”).
 See I.R.C. § 7803(c)(2)(B)(ii)(X) (2012) (mandating that the National Taxpayer Advocate report to Congress annually on the “10 most litigated issues for each category of taxpayers”). Innocent spouse relief has appeared on this “top ten” list every year since 2004. See e.g., Nat’l Taxpayer Advocate, The Most Litigated Issues, FY 2011 Ann. Rep. to Cong. 589 (2011) [hereinafter TAS Report 2011] (listing relief from joint and several liability as the eighth most litigated issue from June 1, 2010, to May 31, 2011). The list includes only those cases for which an opinion is issued; since 2007, more than seventy-five percent of docketed Tax Court cases are settled each year, and fewer than three percent of docketed cases resulted in trial and decision. Id. at 591. Thus, the vast majority of docketed cases are therefore not reflected in these annual lists.
 Once a petition is filed with the Tax Court, cases are automatically sent to an IRS Appeals office, which concentrates on attempting to settle cases prior to trial. That Ms. Bozick’s trial went forward indicates that the parties were unable to do so. More than seventy-five percent of Tax Court cases are settled outside of court. Supra note 20.
 In the 1990s, the recommenders included the American Bar Association, the American Institute of Certified Public Accountants, and other academic commentators. TAS Report 2005, supra note 17, at 409 n.13; see also Dep’t. of the Treas., Report to the Congress on Joint Liability and Innocent Spouse Issues 34–41 (1998) [hereinafter 1998 Treas. Report on JL and IS], available at http://www.treasury.gov/resource-center/tax-policy/Documents/innospos.pdf.
 Id. at 57–58.
 This article will focus primarily on the concerns of low-income taxpayers, but numerous scholars have argued for mandatory individual filing on a variety of policy grounds. See, e.g., Lily Kahng, One is the Loneliest Number: The Single Taxpayer in a Joint Return World, 61 Hastings L.J. 651, 684 (2010) (arguing that the joint return should be abolished because it penalizes single people). See also James Puckett, Rethinking Tax Priorities: Marriage Neutrality, Children, and Contemporary Families, 78 U. Cin. L. Rev. 1409, 1434 (2010) (arguing that “[t]he joint return (and special rates for married taxpayers) should be abolished as an incoherent penalty and subsidy of marriage”); Edward J. McCaffery, Taxing Women 19–23 (1997) (advocating individual filing as more beneficial for women); Lawrence Zelenak, Marriage and the Income Tax, 67 S. Cal. L. Rev. 339 (1994) (concluding that mandatory separate returns present the best option among imperfect neutralities); Marjorie E. Kornhauser, Love, Money, and the IRS: Family, Income Sharing, and the Joint Income Tax Return, 45 Hastings L.J. 63, 108 (1993) (arguing that individual tax returns would be more equitable).
 TAS Report 2005, supra note 17, at 422 (citing data from tax year 2001). For this statistical reason, I refer to the requesting spouse in the feminine or focus my arguments on women in particular. This is not to suggest that men do not file or succeed with claims for innocent spouse relief; in fact, in one of my first innocent spouse cases I represented an ex-husband who was granted relief at the administrative appeal level.
 Most innocent spouse claims are filed by the lower-earning spouse, which is usually the woman. TAS Report 2001, supra note 8, at 132. “Nearly 90% of the claims filed for relief from Joint and Several Liability are filed by women with earned income that is approximately 25% of the total income on a joint return.” Id. at 128.
 Congress funds a Low-Income Taxpayer Clinic (LITC) program administered by the Taxpayer Advocate Service. This program, which was established in 1998 with the addition of I.R.C. § 7526, provides federal matching grants to non-profit organizations that provide legal representation to income-eligible taxpayers who have a controversy with the Internal Revenue Service. I direct the LITC program at Washington and Lee University School of Law, and this article is inspired in large part by my experiences working with these clients.
 Supra note 17.
 See The Nat’l Marriage Project, The State of Our Unions: The Social Health of Marriage in America 29 (2003) (charting the 850% increase of unmarried, cohabiting, adult couples living with one or more children under age 15 from the 1960s to 2000; the figure increased from 197,000 in 1960 to 1.675 million in 2000).
 For a detailed history of joint and several liability, see 1998 Treas. Report on JL and IS, supra note 22. See also Richard C. E. Beck, The Innocent Spouse Problem: Joint and Several Liability for Income Taxes Should Be Repealed, 43 Vand. L. Rev. 317 (1990) (providing a detailed history of the origins of the joint return, Cole, and the subsequent introduction of joint liability); Bryan T. Camp, The Unhappy Marriage of Law and Equity in Joint Return Liability, 108 Tax Notes 1307 (2005).
 Revenue Act of 1913, Pub. L. No. 63-16, ch. 16, § II (A)(1), 38 Stat. 114, 166 (1913).
 Revenue Act of 1916, Pub. L. No. 64-271, ch. 463, § 1(a), 39 Stat. 756 (1916).
 See Beck, supra note 30, at 335; see also TAS Report 2001, supra note 8, at 132. Separate rate brackets for joint returns were not introduced until 1948.
 I.T. 1575, 1923-1 C.B. 144 (1923) (“[W]here a husband and wife filed a joint return they are individually liable for the full amount of tax shown to be due on such return.”).
 Helvering v. N.Y. Trust Co., 292 U.S. 455 (1934).
 See Cole v. Comm’r, 81 F.2d 485, 488 (9th Cir. 1935) (noting the conflict between the Commissioner’s arguments and the cautionary notice in the bulletin).
 Id. (citing Helvering, 292 U.S. at 467–68). See also Comm’r v. Rabenold, 108 F.2d 639, 641 (2d Cir. 1940) (“[O]nly by implication can such a liability be found in the 1932 Act, and taxing laws are not to be extended by implication.”).
 Cole, 81 F.2d. at 489.
 Id. at 486.
 Id. at 490.
 Id. (citing Gummey v. Comm’r, 26 B.T.A. 894, 895–96 (1932)).
 Id. at 487 (emphasis added).
 See, e.g., Uniacke v. Comm’r, 132 F.2d 781, 783 (2d Cir. 1942) (ruling that joint and several liability did not attach). In a case involving tax year 1936, the court stated: “The argument based on administrative convenience does not impress us . . . . The administrative difficulties resulting from rejecting the Commissioner’s construction of the statute would seem to be somewhat exaggerated in the argument.” See also Estate of Hague v. Comm’r, 45 B.T.A. 104, 113 (1941) (in which the Commissioner argued for joint and several liability for a couple where the income in question was solely the husband’s); Rabenold, 108 F.2d at 640.
 Revenue Act of 1938, Pub. L. No 75-554, ch. 289, § 51(b), 52 Stat. 447, 476 (1938). The Internal Revenue Code of 1986, as amended, retains the same language in § 6013(d)(3): “if a joint return is made, the tax shall be computed on the aggregate income and the liability with respect to the tax shall be joint and several.”
 Rabenold, 108 F.2d at 640.
 See Grant v. Rose, 24 F.2d 115, 118–19 (N.D. Ga. 1928) (summarizing these advantages and disadvantages), aff’d, 39 F.2d 340 (5th Cir. 1930).
 Revenue Act of 1948, Pub. L. No. 80-471, ch. 168, § 301, 62 Stat. 110, 114 (1948).
 Poe v. Seaborn, 282 U.S. 101, 118 (1930).
 State laws in community property states vary as to whether the interests of each spouse are subject to lien or levy by the Service. In at least some community property states, the Service can collect the separate tax liability of one spouse from the other spouse’s community income. See 1998 Treas. Report on JL and IS, supra note 22. In 1980, Congress enacted § 66 to provide limited relief to taxpayers in community property states. Section 66, while important to those who live in community property states, is outside the scope of this article.
 See S. Rep. No. 80-1013, at 22 (1948), reprinted in 1948 U.S.C.C.A.N. 1163, 1184 (describing the existing law: “Since the rates applied under the income tax are steeply progressive, the same family income divided in two halves by community property law will be taxed far less severely than in a common-law State where the whole income is apt to be taxed to one spouse.”).
 Id., at 24–25. The legislative history gives another interesting reason for adopting income splitting in the Code: “The incentive for married couples in common-law states to attempt the reduction of their taxes by the division of their income through such devices as trusts, joint tenancies, and family partnerships will be reduced materially. Administrative difficulties stemming from the use of such devices will be diminished, and there will be less need for meticulous legislation on the income-tax treatment of trusts and family partnerships.”
 Revenue Act of 1948, Pub. L. No. 80-471, ch. 168, § 301, 26 U.S.C. §12, 62 Stat. 110, 114 (1948).
 See Beck, supra note 30, at 337 n.78 (citing legislative history showing that, in 1938, when it was rarely the case that a couple would benefit economically by filing jointly, ninety-four percent of married couples filed a joint return) (citing H.R. Rep. No. 1040, 77th Cong., 1st Sess. at 16-17 (1941), reprinted in 1941-2 C.B. 413, 426-27). Beck points out that federal income tax rates were low enough in 1938 that it made no difference in most people’s tax rates whether they filed jointly or separately; he posits that most of these couples filed jointly for the sake of the convenience of filling out one form rather than two, rather than to achieve tax savings.
 For a historical examination of marriage bonuses and penalties, see Lawrence Zelenak, Doing Something About Marriage Penalties: A Guide for the Perplexed, 54 Tax L. Rev. 1 (2000). Zelenak notes that while the 1948 legislation achieved couples neutrality, its purpose was “a delayed response to geographic discrimination between husbands in separate property states and those in community property states.” Id. at 4. Zelenak notes, as others have, that the goals of progressivity, couples neutrality, and marriage neutrality are incompatible. In both this article and in his article entitled Marriage and the Income Tax, supra note 24, Zelenak states his preference for mandatory separate filing.
 Staff of Joint Comm. on Taxation, 99th Cong., General Explanation Of 1969 Tax Reform Act 222 (Joint Comm. Print 1970).
 Id. at 223.
 In his 1990 article, Beck said the following about the introduction of the marriage penalty in the Tax Reform Act of 1969: “The argument that income splitting justifies joint return liability seems hollow for that forty percent of married couples who suffer the marriage penalty. While they may be better off filing jointly than filing separately, they would be still better off if they had never married, or were divorced. From this vantage point, filing jointly, as ninety-nine percent of married persons do, involves an overall tax detriment, rather than a benefit. If joint return liability were justified on a quid pro quo theory, the liability probably should have been abolished in 1969, when the marriage penalty arose.” Beck, supra note 30, at 372.
 I conclude it is appropriate to reconsider the current structures in light of other code provisions and changing demographics. See infra Part IV.
 The Nat’l Marriage Project, Executive Summary to Why Marriage Matters: Thirty Conclusions from the Social Sciences 1 (3d ed. 2011), available at http://www.americanvalues.org/pdfs/dl.php?name=wmm3es.
 The Treasury Department invoked this rationale in its 1998 report eschewing separate filing in favor of expanded innocent spouse reform. See infra Part III. But see Comm’r v. Uniacke, 132 F.2d 781, 783 (2d Cir. 1942) (dismissing the rationale of administrative convenience); Comm’r v. Rabenold, 108 F.2d 639, 640 (2d Cir. 1940) (recognizing the administrative inconvenience of apportioning tax deficiencies between husband and wife). Both cases were decided under pre-1938 law.
 1998 Treas. Report on JL and IS, supra note 22, at 6–7 (citing H.R. 1860, 75th Cong. 2d Sess., 1939-1 C.B. 749) (emphasis added).
 Id. at 25–27.
 Internal Rev. Serv., One Billion Served: IRS E-File Passes Major Milestone, IR-2011-64 (June 9, 2011), available at http://www.irs.gov/uac/One-Billion-Served:-IRS-E-File-Passes-Major-Milestone.
 Nat’l Taxpayer Advocate, Key Legislative Recommendation: Another Marriage Penalty – Taxing the Wrong Spouse, 2005 Ann. Rep. to Cong. 426 (citing IRS Pub. 55B, Data Book (Mar. 1998) (Tables 3 and 4)).
 Id. at 408–09.
 See infra Part III.
 Nat’l Taxpayer Advocate, supra note 64, at 425–26.
 Id. at 426. Olson mentions community property relief, injured spouse relief, and “numerous other IRC provisions that override community property rules in an ad hoc fashion.” Id. at 426. While outside the scope of this article, each of these issues exists because of joint and several liability.
 Internal Rev. Serv., Taxpayers Who File Electronically Must Use e-Signatures, FS-2011-07 (Jan. 2011), available at http://www.irs.gov/pub/irs-news/fs-11-07.pdf.
 I.R.M. § 188.8.131.52.4 (Mar. 4, 2011) available at http://www.irs.gov/irm/part25/irm_25-015-001.html#d0e231 (providing guidance to Service employees as to the conditions under which a joint return is invalid). These questions are technically outside of innocent spouse relief, because § 6015 relief is only available when the requesting spouse has signed a joint income tax return.
 Cole v. Comm’r, 81 F.2d 485, 487 (9th Cir. 1935).
 H.R. Rep. No. 1860, at 29–30 (1938).
 Sonnenborn v. Comm’r, 57 T.C. 373, 380–81 (1971) (discussing why the innocent spouse provisions must be narrowly construed) (emphasis added). Section 6013(e) was the innocent spouse provision in effect at the time. As the opinion notes in footnote 3, the innocent spouse provision became law on January 12, 1971, but by virtue of its enacting legislation the relief was applicable to all taxable years to which the Internal Revenue Code of 1954 applied.
 See, e.g., Ogonoski v. Comm’r, 87 T.C.M. (CCH) 1038; Crowley v. Comm’r, 70 T.C.M. (CCH) 1374; Prince v. Comm’r, 70 T.C.M. (CCH) 309; Murphy v. Comm’r, 103 T.C. 111 (1994).
 Sonnenborn did not use the term “price” with respect to the privilege, but subsequent courts citing Sonnenborn for this principle have used this term. See, e.g., Stevens v. Comm’r, 872 F.2d 1499, 1503 (11th Cir. 1989) (“The rate of tax applied against a given amount of income generally is lower when the income is reported on a joint return than when a husband and wife file separate returns. The price which the law exacts for this privilege is that taxpayers who file a joint return are jointly and severally liable for the amount of tax due, regardless of the source of income reported and notwithstanding the fact that one spouse may be less informed about the contents of the return.”) (citation omitted), aff’g 55 T.C.M. (CCH) 135; Estate of Woodward v. Comm’r, T.C. Memo 1995-523 (“The privilege of filing a joint income tax return does not come without a price, however. Thus, as a general rule, spouses who file joint income tax returns are jointly and severally liable for the full amount of tax due on the combined incomes.”).
 As mentioned in note 4, supra, in almost all cases married taxpayers face a higher liability if they elect the married filing separately status. This holds true for two high-earners, who face a penalty on a married filing joint return relative to what two unmarried taxpayers would pay in the aggregate yet would be subject to an even higher liability if they elected married filing separately status.
 See e.g., Dranow v. Comm’r, 27 T.C.M. (CCH) 1485 (1968). Forgery is to be distinguished from signing someone else’s name with tacit consent. Thus if one spouse signs for the other, the court will uphold joint liability where it finds that the non-signing spouse tacitly consented to the signing spouse’s action. See, e.g., Abrams v. Comm’r, 53 T.C. 230 (1969).
 See e.g., Stanley v. Comm’r, 45 T.C. 555, 563 (1966) (holding that the taxpayer “failed to prove the necessary causal relationship between her fear of [her husband] and her signing of the returns”); Federbush v. Comm’r, 34 T.C. 740, 754–58 (1960), aff’d per curiam, 325 F.2d 1 (2d Cir. 1963); Estate of Aylesworth v. Comm’r, 24 T.C. 134, 145–46 (1955) (holding that taxpayer’s abuse at the hands of her husband did not constitute duress). The Aylesworth opinion noted: “The filing of joint returns resulted in a substantially reduced tax burden for the couple as a result of the split income provisions, and we should be very slow to conclude that the signature of the wife is to be regarded as having been obtained by fraud or duress.” Id. at 146.
 Scudder v. Comm’r, 48 T.C. 36, 41 (1967). Mrs. Scudder appealed to the Sixth Circuit Court of Appeals, which disagreed with the Tax Court that nothing could be done. It remanded the case for further consideration of the innocence of the fraud as a defense. Scudder v. Comm’r, 405 F.2d 222 (6th Cir. 1968). Another oft-cited case that included a plea for a legislative fix is Wissing v. Comm’r, 54 T.C. 1428, 1432 (1970) (“We would welcome a rule which would grant relief to a victimized spouse who has no knowledge of or reason to have knowledge of, and does not benefit from, unreported income, at least where that income is the fruit of a crime. But we regretfully see no way in which this Court can or should engraft such a ‘doing equity’ rule on the language of section 6013(d)(3). We think that such a result should properly be accomplished by ameliorating legislation.”), vacated, 441 F.2d 533 (6th Cir. 1971).
 Innocent Spouse Act of 1971, Pub. L. No. 91-679, 84 Stat. 2063. In 1998, the innocent spouse provisions were amended and moved to § 6015, where they remain today.
 Staff of Joint Comm. on Tax’n, 105th Cong., Present Law and Background Relating to Tax Treatment of “Innocent Spouses” (Joint Comm. Print 1998) (citing S. Rep. No. 91-1537, at 2 (1970)).
 S. Rep. No. 91-1537, at 2 (1970).
 I.R.C. § 6015(f)(1) (2012).
 H.R. Rep. No. 104-506, at 30 (1996).
 Taxpayer Bill of Rights 2, Pub. L. No. 104-168, § 401 (1996).
 H.R. Rep. No. 104-506, at 31 (1996).
 I.R.S. Notice 96-19, 1996-1 C.B. 371.
 See generally 1998 Treas. Report on JL & IS, supra note 22.
 Id. at 57 (quoting Taxpayer Bill of Rights 2, supra note 86).
 See id. at 58. Part III will examine the Treasury Department’s arguments in favor of maintaining the joint and several liability standard.
 See IRS Restructuring: Hearings on H.R. 2676 Before the S. Comm. on Fin.,105th Cong. (1998).
 IRS Restructuring and Reform Act of 1998, H.R. 2676, 105th Cong. § 3201 (1998) (enacted) [hereinafter 1998 Act]; I.R.C. § 6015 (2012).
 The final version of the bill did not allow for separation of liability for underpayments, as had been proposed by the Senate amendment. However, relief for underpayments was made available in § 6015(f).
 See Svetlana Attestatova, The Bonds of Joint Tax Liability Should Not Be Stronger than Marriage: Congressional Intent Behind § 6015(c) Separation of Liability Relief, 78 Wash. L. Rev. 831, 858 (2003) (including an in-depth examination of the meaning of “any item”).
 Prior to 1998, duress was (and still is) a common law defense to joint and several liability. See supra note 78.
 I.R.C. § 6015(f)(1) (2012).
 I.R.C. § 6015(f)(2) (2012).
 H.R. Rep. No. 105-599, at 254 (1998) (Conf. Rep.).
 The current guidelines are set out in Rev. Proc. 2003-61, 2003-2 C.B. 296, which superseded Revenue Procedure 2000-15, 2000-1 C.B. 447. Notice 2012-8, 2012-4 I.R.B. 309, introduced a proposed revenue procedure that, if adopted, will supersede Revenue Procedure 2003-61. See also Treas. Reg. § 1.6015-5 (2002).
 H.R. Rep. No. 105-599, at 254 (1998) (Conf. Rep.).
 The first version of IRS Form 8857, Request for Innocent Spouse Relief (And Separation of Liability and Equitable Relief), was released in December 1998 and was a one page form with three pages of instructions. The form has been revised five times since then and is currently under revision again. The most recent version of the form (Rev. September 2010) is four pages long and includes five pages of separate instructions. All versions are available at http://www.irs.gov/uac/Form-8857,-Request-for-Innocent-Spouse-Relief.
 The Cincinnati Centralized Innocent Spouse Operation (CCISO) is located in Covington, Kentucky, and was established in fiscal year 2001. U.S. Gov’t Accountability Office, GAO-02-558, Innocent Spouse Program Performance Improved; Balanced Performance Measures Needed 4 (2002).
 See, e.g., Nat’l Taxpayer Advocate, The Most Serious Problems Encountered By Taxpayers, FY 2001 Ann. Rep. to Cong. 67–68. IRS Publication 971, Innocent Spouse Relief, was originally released in 1998 and revised six times since. The current version (Rev. September 2011) is 24 pages long. The instructions to Form 1040 (Tax Year 2011) include an explanation of joint and several liability and information about Innocent Spouse Relief; however, at 189 pages, it cannot be assured that taxpayers will read this information in the instructions at the time of filing. The Service developed an online Innocent Spouse Tax Relief Eligibility Explorer, which is available on the IRS website at http://www.irs.gov/Individuals/Explore-if-you-are-an-Eligible-Innocent-Spouse.
 See supra note 19.
 Rev. Proc. 2003-61, 2003-2 C.B. 296.
 Id. at § 4.01.
 Id. at § 4.02.
 Id. at § 4.03 (providing a nonexclusive list of eight factors that the Service will consider in making its determination).
 See supra note 9.
 Bozick v. Comm’r, 99 T.C.M. (CCH) 1242, 1244 (2010).
 See, e.g., Fernandez v. Comm’r, 114 T.C. 324, 332 (2000); Ewing v. Comm’r, 122 T.C. 32, 36–38 (2004).
 Comm’r v. Ewing, 439 F.3d 1009, 1014 (9th Cir. 2006); see also Bartman v. Comm’r, 446 F.3d 785, 787–88 (8th Cir. 2006) (holding that the Tax Court lacked jurisdiction under § 6015(e)(1) where there was no deficiency).
 I.R.C. § 6015(e)(1)(A) (2012).
 Ewing v. Comm’r, 122 T.C. 32 (2004), vacated, 439 F.3d 1009 (9th Cir. 2006). In Porter v. Comm’r, 130 T.C. 115, 124 (2008) [hereinafter Porter I], the Tax Court revisited the issue and upheld the position that the court may consider evidence introduced at trial even if not included in the administrative record. Porter I addressed the scope of review question, but did not resolve the standard of review question; it stated that determination of the scope of review did not depend on the standard of review applied. Porter I, 130 T.C. at 122 n.10.
 See Porter v. Comm’r, 132 T.C. 203 (2009) [hereinafter Porter II] (revisiting the standard of review question in § 6015(f) cases). Following the 1998 Act, the court used the de novo standard of review in §§ 6015(b) and (c) cases, but held in Butler v. Comm’r, 114 T.C. 276, 292 (2000) that abuse of discretion was the appropriate standard of review in § 6015(f) equitable relief cases. In Porter II, the court revisited this question in light of Congress’s 2006 amendments to § 6015(e)(1); it concluded that by the language of the amendments, Congress intended the court to use a de novo scope of review and standard of review in cases determining relief under § 6015(f). Porter II, 132 T.C. at 208. But see the dissent in Porter II, arguing that Congress did not intend in its 2006 amendment to § 6015(e) to change the court’s standard of review from abuse of discretion to de novo. Porter II, 132 T.C. at 233–35 (Gustafson, J., dissenting).
 Nat’l Taxpayer Advocate, Legislative Recommendations 1, 2011 Ann. Rep. to Cong. 534–35 (2011) (citing as an example Torrisi v. Comm’r, T.C.M. 2011-235, slip op. at 17 n.15). The Taxpayer Advocate called upon Congress to amend § 6015 to clarify that the Tax Court’s scope and standard of review is de novo in § 6015(f) cases, stating: “[t]he IRS’s position is especially harmful to taxpayers who cannot afford representation or assistance during administrative proceedings, or those who are victims of domestic violence or abuse. The divergence between Counsel’s position and that of the Tax Court creates uncertainty for taxpayers and consumes administrative and judicial resources.” Id. at 474.
 I.R.S. Notice CC-2009-021 (June 30, 2009).
 I.R.C. § 6015(h)(2) (2012) provides the nonrequesting spouse the statutory right to respond. While an important procedural safeguard, the process puts the Service into the position of fact-finder in a “he-said, she-said” argument between ex-spouses. Should the requesting spouse prevail, the nonrequesting spouse has the right to appeal. Rev. Proc. 2003-19, 2003-1 C.B. 371.
 Nat’l Taxpayer Advocate, Most Litigated Issues 2, 2011 Ann. Rep. to. Cong. 659 (2011).
 Lantz v. Comm’r, 607 F.3d 479 (7th Cir. 2010).
 See I.R.C. § 6015 (2012). “Collection activity” is defined in Treas. Reg. § 1.6015-5(b)(2)(i) (2002) to mean: “a section 6330 notice; an offset of an overpayment of the requesting spouse against a liability under section 6402; the filing of a suit by the United States against the requesting spouse for the collection of the joint tax liability; or the filing of a claim by the United States in a court proceeding in which the requesting spouse is a party or which involves property of the requesting spouse.”
 Prior to the promulgation of the regulation, the Service issued several forms of public guidance announcing its position that the two-year limitation applied to § 6015(f) claims. See I.R.S. Notice 98-61, § 3.01(3), 1998-2 C.B. 758; Rev. Proc. 2000-15, 2000-1 C.B. 447; Rev. Proc. 2003-61, 2003-2 C.B. 296.
 Lantz v. Comm’r, 132 T.C. 131, 150 (2009), rev’d, 607 F.3d 479 (7th Cir. 2010).
 Lantz v. Comm’r, 607 F.3d 479 (7th Cir. 2010).
 The Tax Court follows the precedent of a Court of Appeals in subsequent cases that would be appealed to that particular circuit; as a result, it often rules differently according to where the appeal would lie. See Golsen v. Comm’r, 54 T.C. 742, 756–57 (1970), aff’d, 445 F.2d 985 (10th Cir. 1971).
 The Third and Fourth Circuits followed the Seventh Circuit in upholding the regulation. See Mannella v. Comm’r, 631 F.3d 115, 121–25 (3d Cir. 2011); Jones v. Comm’r, 642 F.3d 459, 465 (4th Cir. 2011). The Service filed appeals in the Second, Sixth, and Ninth Circuits, but later dropped these appeals in light of I.R.S. Notice 2011-70, 2011-32 I.R.B. 135 (infra note 136). See Coulter v. Comm’r, T.C. No. 1003-09 (stipulated decision Nov. 20, 2009), appeal docketed, No. 10-680 (2nd Cir. Feb. 24, 2010); Buckner v. Comm’r, T.C. No. 12153-09 (decided May 21, 2010), appeal docketed, No. 10-2056 (6th Cir. Aug. 18, 2010); and Carlile v. Comm’r, T.C. No. 011567-09 (decided May 26, 2010), appeal docketed, No. 10-72578 (9th Cir. Aug. 23, 2010). Despite the appeals, the Tax Court continued to hold the two-year deadline invalid in the circuits that had not yet ruled. See, e.g., Pullins v. Comm’r, 136 T.C. 432, 441–42 (2011); Kelly v. Comm’r, 100 T.C.M. (CCH) 507, 510 (2010); Hall v. Comm’r, 135 T.C. 374, 382 (2010). Pullins was on appeal in the Eighth Circuit, and both Kelly and Hall were on appeal in the Sixth Circuit; these Tax Court opinions were filed subsequent to the Buckner opinion but prior to I.R.S. Notice 2011-70.
 See, e.g., Patrick J. Smith, Gaps in the Seventh Circuit’s Reasoning in Lantz, 2010 Tax Notes Today 186-19 (2010) (criticizing the Seventh Circuit opinion); Bryan T. Camp, Interpreting Statutory Silence, 2010 Tax Notes Today 148-6 (2010) (analyzing the Tax Court’s decision).
 See Nat’l Taxpayer Advocate, Allow Taxpayers to Request Equitable Relief Under Internal Revenue Code 6015(f) or 66(c) at Any Time Before Expiration of the Period of Limitations on Collection and to Raise Innocent Spouse Relief as a Defense in Collection Actions 1, 2010 Ann. Rep. to Cong. 377–78 (2010).
 Press Release, S. Fin. Comm., Baucus, Harkin, Sherrod Brown Call on IRS to Give Innocent Spouses More Time to File For Tax Relief (Apr. 18, 2011).
 I.R.S. Notice 2011-70, 2011-32 I.R.B. 135.
 I.R.S. News Release IR-2011-80 (July 25, 2011).
 Nat’l Taxpayer Advocate, Unlimit Innocent Spouse Equitable Relief, 2 2010 Ann. Rep. to Cong. 11 at n.54 (2010).
 I.R.S. Notice 2012-8, 2012-4 I.R.B. 309.
 I.R.S. News Release IR-2012-03 (Jan. 5, 2012).
 See supra note 139, at § 4.02.
 See id. at § 4.03(2)(c)(ii).
 See supra notes 93 and 135.
 Taxpayer Bill of Rights 2, Pub. L. No. 104-168, § 401, 110 Stat. 1452, 1459 (1996).
 1998 Treas. Report on JL and IS, supra note 22, at 2.
 Id. at 3.
 Id. at 3. The report also addressed issues related to community property laws and the rule of Poe v. Seaborn. I have largely ignored the effect of community property laws in this article because I presume that if Congress mandated individual filing, it would also legislatively overrule Poe v. Seaborn. Along with a chorus of scholars, Taxpayer Advocate Nina Olson has called for the legislative repeal of Poe v. Seaborn. Nat’l Taxpayer Advocate, Key Legislative Recommendation: Another Marriage Penalty—Taxing The Wrong Spouse 1, 2005 ANN. REP. TO CONG. 407, 409 (2005).
 Id. at 25 n.38.
 Id. at 25–26.
 Id at 26.
 See supra notes 63 and 64.
 1998 Treas. Report on JL and IS, supra note 22, at 34. The Report notes that the ABA proposal represented the views of individuals who submitted the comments and did not represent an official position of the ABA or its Tax Section.
 Id. at 38–39.
 Id. at 38, 41.
 Id. at 42 (citing H.R. Rep. No. 104-56, at 30 (1996)).
 See supra Part I.B.1; supra note 63.
 See, e.g., Beck, infra note 161, at 950–51, 954.
 Treas. Inspector Gen. for Tax Admin., The Process to Separate Joint Tax Accounts for Innocent Spouse Cases Has Been Improved; However, Additional Actions Are Needed (2007), available at http://www.treasury.gov/tigta/auditreports/2007reports/200740053fr.pdf.
 Id. at 4.
 See Richard C.E. Beck, The Innocent Spouse Problem: Joint and Several Liability for Income Taxes Should Be Repealed, 43 Vand. L. Rev. 317 (1990).
 See Richard C.E. Beck, The Failure of Innocent Spouse Reform, 51 N.Y.L. Sch. L. Rev. 929 (2006–07).
 Id. at 950.
 Id. at 950–51 (citing TAS Report 2005, supra note 17, at 423).
 Id. at 954.
 Id. at 932.
 Id. at 954.
 Stephanie Hunter McMahon, An Empirical Study of Innocent Spouse Relief: Do Courts Implement Congress’s Legislative Intent?, 12 Fla. Tax Rev. 629 (2012).
 Id. at 706.
 Id. (quoting Beck, supra note 161, at 942).
 Id. at 707.
 1998 Treas. Report on JL and IS, supra note 22, at 3.
 Boris Bittker examined the Code’s treatment of married couples as an economic unit and framed the tensions of the competing neutralities in his well-known 1975 essay, Federal Income Taxation and the Family, 27 Stan. L. Rev. 1389. A number of scholars have since considered these issues and concluded that, notwithstanding the neutrality trade-offs, it would be more equitable to treat each married individual as a separate economic unit. See, e.g., Kornhauser, supra note 24, at 108; Zelenak, supra note 24; Beck, supra note 160, at 382 (arguing that even if couples do pool their resources, this “cannot justify any transfer of tax liability from one spouse to the other”).
 1998 Treas. Report on JL and IS, supra note 22, at 41–44.
 This solution presumes that family law judges and practitioners would be consistent about including this form in the divorce proceedings. As a general proposition, low-income clients are less likely to have adequate or comprehensive representation, if any, during divorce proceedings.
 The 1998 Treasury Report on Joint Liability and Innocent Spouse Issues raised a concern that some form of equitable relief might still be necessary if this proposal were adopted, because a tax liability might not arise until after the divorce was finalized. See 1998 Treas. Report on JL and IS, supra note 22, at 44. This is true, but the possibility could be addressed within the proposal. The spouses are in the best position to know which individual benefits most from joint filing, as well as which individual’s return items are more likely to be scrutinized under audit. The standard IRS form could stipulate that if it were subsequently discovered that one spouse underreported his or her income, that individual is wholly responsible for any corresponding adjustment in tax. It does not seem unreasonable for one spouse to accept the risk of a subsequent audit or adjustment to liability as part of a divorce settlement; alternatively, the spouses could agree on the form to divide the risk by allocating the tax responsibility evenly.
 The Nat’l Comm’n on Fiscal Responsibility and Reform, The Moment of Truth, available at http://www.fiscalcommission.gov/sites/fiscalcommission.gov/files/documents/TheMomentofTruth12_1_2010.pdf; see also Bipartisan Policy Ctr., Restoring America’s Future: Reviving the Economy, Cutting Spending and Debt, and Creating a Simple, Pro-Growth Tax System, available at http://bipartisanpolicy.org/sites/default/files/BPC%20FINAL%20REPORT%20FOR%20PRINTER%2002%2028%2011.pdf.
 Kornhauser, supra note 24, at 66-68 (discussing the “difficulty in defining ‘family’” given the demographic changes in the thirty year period preceding her article).
 Id. at 66 (citing U.S. Bureau of the Census, Statistical Abstract of the United States, Table 56 (112th ed. 1992)).
 Id. at 80–91 (discussing “empirical studies of asset pooling and control sharing”).
 Id. at 105.
 Id. at 65 (“Taxation based on the individual comports better with reality, social policies promoting families, tax theory, and economic considerations than the joint return.”).
 D’Vera Cohn et al., Pew Research Ctr., New Marriages Down 5% from 2009 to 2010: Barely Half of U.S. Adults Are Married—A Record Low 1 (2011), available at http://www.pewsocialtrends.org/files/2011/12/Marriage-Decline.pdf.
 Id. at 10. Interestingly, 47% of the unmarried adults who said that marriage is becoming obsolete also said that they would like to get married someday. Id. at 11.
 Id. at 10.
 Id. at 8.
 See Kahng, supra note 24.
 Anthony Infanti, Decentralizing Family: An Inclusive Proposal for Individual Tax Filing in the United States, 3 Utah L. Rev. 605, 607–08 (2010).
 One might wonder why a spouse would be at all concerned with joint and several liability if the couple is to receive a large earned income credit. After all, if the couple will receive a credit, there is no liability to attach. But this is not always the reality. Consider a husband who works as an independent contractor, meaning he is not subject to withholding. Assume that he and his wife have children, and that his wife either does not work or is a wage-earner who is subject to proper withholding. Without the benefit of the earned income credit and the child tax credit (the latter not being contingent on filing jointly), the couple would have a significant tax liability due to the husband’s absence of withholding, failure to make estimated tax payments, and self-employment tax liability. Even with the benefits of such credits, the couple may face a tax liability, but it would be far smaller. Thus, the couple is dependent on the earned income credit to reduce what would otherwise be a liability that they likely could not afford.
 And I do not.
 TAS Report 2005, supra note 25.
 All figures here ignore the Making Work Pay credit, enacted as part of the American Recovery and Reinvestment Act of 2009, which provided a refundable credit of up to $400 per working taxpayer in tax years 2009 and 2010.
 The calculations assumed that neither taxpayer had any federal tax withholdings. The taxpayers’ preliminary joint tax liability of $703 was eliminated by the child tax credit, meaning no tax was owed. The resulting $3,897 refund is attributable to $1,297 for the additional child tax credit and $2,600 for the earned income credit.
 His preliminary liability of $838 is offset by the child tax credit. His refund is attributable the additional child tax credit of $1,162. Though he would be income eligible if he were unmarried, he will not receive any earned income credit because I.R.C. § 32(d) (2012) disallows it for “married filing separately” returns.
 She has no preliminary liability due to her standard deduction and personal exemption, and she will not receive any refundable credits.
 Her adjusted gross income was $8,000. Her standard deduction of $5,700 plus her personal exemption of $3,650 equals a total of $9,350. However, one cannot have a negative taxable income. Thus, she could have earned an additional $1,350 before being subject to any tax liability.
 It would be rational for the partners to allocate the children to the higher earner, because the lower earner benefits considerably less from the corresponding dependency exemptions and child tax credit.
 The male partner’s preliminary liability of $838 would be eliminated by the child tax credit; his refund of $4,392 consists of a $3,230 earned income credit and a $1,162 additional child tax credit.
 The female partner’s preliminary liability was $0; her refund is attributable to a $416 earned income credit.
 Note that this is not this is not always the case: as the earned income credit phases out according to adjusted gross income, it phases out at a lower adjusted gross income level for unmarried filers than for married filing jointly. For example, in 2010, a head of household filer with two children and an adjusted gross income of $42,000 will receive no earned income credit, while a married filing jointly couple with two children and a combined adjusted gross income of $42,000 will receive a $705 earned income credit. The married filing jointly couple will receive no earned income credit once their adjusted gross income reaches $45,373.
 See I.R.C. § 2(b) (2012). The father will be eligible because he is unmarried, maintains a household with a qualifying child, and furnishes more than half the cost of maintaining the household. The head of household status was enacted in 1951 as a way to extend tax relief to single persons with dependents: “The income of a head of household who must maintain a home for a child, for example, is likely to be shared with the child to the extent necessary to maintain the home, and raise and educate the child. This, it is believed, justifies the extension of some of the benefits of income splitting. The hardship appears particularly severe in the case of the individual with children to raise who, upon the death of his spouse, finds himself in the position not only of being denied the spouse’s aid in raising the children, but under present law also may find his tax load heavier.” H.R. Rep. No. 82-586, at 11 (1951). Extrapolating from this Congressional intent, it is somewhat illogical then that a two-income unmarried couple raising children together should benefit from a provision intended for single parents. But there is no provision limiting the status to those who maintain a household without support from another adult. This is a further and arguably more egregious example of the “unmarriage bonus” and a result that flies in the face of the policy intended by Congress.
 The father’s rate remains the same regardless of filing status, but the standard deduction for “head of household” in 2010 was $8,400, while the standard deduction for “single” was only $5,700. Thus, his taxable income is $2,700 less in Scenario 4 than in Scenario 3. In turn, this lower taxable income reduces his preliminary tax liability by $270, which creates a corresponding $270 increase in his refundable additional child tax credit. The father’s preliminary tax of $568 is eliminated by the child tax credit, and his $4,662 refund consists of a $3,230 earned income credit and a $1,432 additional child tax credit.
 In the case of a married couple, both spouses’ incomes must be considered when determining earned income credit eligibility. If, however, the spouses’ aggregate income is within the income limits, it does not seem fair to deny both spouses the earned income credit for filing separately. However, a problem may arise later if one spouse is determined to have understated his or her income. If in fact the husband earned not $25,000 but $50,000, then neither spouse should be eligible under current law. If married taxpayers filing separately were allowed the earned income credit, but were later ineligible due to underreporting by one spouse, it would affect both spouses’ earned income credit eligibility. Should the underreporting spouse bear the sole responsibility, including the liability arising from the other spouse’s incorrect receipt of the earned income credit?
 See generally H.R. Rep. No. 94-19 (1975).
 Tax Reduction Act of 1975, Pub. L. No. 94-12, § 209(b), 89 Stat. 26, 35.
 Joint Comm. on Internal Revenue Tax’n, 94th Cong., Analysis of the House version of the Tax reduction act of 1975 (H.R. 2166) and possible alternatives 33 (Comm. Print 1975). See also S. Rep. No. 94-36, at 11 (1975) (elaborating on the earned income credit as follows: “This new refundable credit will provide relief to families who currently pay little or no income tax. These people have been hurt the most by rising food and energy costs. Also, in almost all cases, they are subject to the social security payroll tax on their earnings. Because it will increase their after-tax earnings, the new credit, in effect, provides an added bonus or incentive for low-income people to work, and therefore, should be of importance in inducing individuals with families receiving Federal assistance to support themselves. Moreover, the refundable credit is expected to be effective in stimulating the economy because the low-income people are expected to spend a large fraction of their increased disposable incomes.”).
 Tax Reduction Act of 1975 § 204. As originally enacted, it appeared in § 43 of the Code. The earned income credit is currently found in I.R.C. § 32 (2012).
 I.R.C. § 32(d) (2012). The legislative history does not provide a clear rationale for denying a credit to married couples filing separately. One possibility is that Congress wanted to keep the credit simple to administer. H.R. Rep. No. 94-19, at 30–31 (1975) provides: “The credit is to be calculated on a return-by-return basis. Individuals who are married and filing a joint return are eligible for only one credit on the combined income of both individuals. Married individuals filing separate returns are not eligible for the credit.” The Senate Report 94-36 contains nearly identical language with no further elaboration. S. Rep. No. 94-36, at 35 (1975).
 The provision that increased the earned income credit for a taxpayer with three or more children was enacted as part of the American Recovery and Reinvestment Act of 2009, Pub. L. No. 111-5, 123 Stat. 115. It was intended as a temporary measure only for tax years 2009 and 2010; however, it was extended through 2012 by the Tax Relief, Unemployment Insurance Reauthorization and Job Creation Act of 2010, Pub. L. No. 111-312, 124 Stat. 3296.
 Jimmy Charite, Indivar Dutta-Gupta & Chuck Marr, Ctr. On Budget & Policy Priorities, Studies Show Earned Income Credit Encourages Work and Success in School and Reduces Poverty (2012) available at http://www.cbpp.org/files/6-26-12tax.pdf.
 Gordon B. Dahl & Lance Lochner, Nat’l Bureau of Econ. Research, The Impact of Family Income on Child Achievement: Evidence from the Earned Income Credit (2011) available at http://dss.ucsd.edu/~gdahl/papers/children-and-EITC.pdf. The authors conclude from their baseline estimates that a $1,000 increase in income raises a child’s contemporaneous math and reading test scores by 6% of a standard deviation. They conclude that their estimates “suggest that the effects are larger for children growing up in more disadvantaged families, younger children, and boys.” Id. at 1951. See also Charite, Dutta-Gupta & Marr, supra note 217, at 2, 8–10 (citing a number of different studies in support of the report’s statement that “[a] small but growing body of research indicates that lifting the incomes of low-income families helps children in those families do better in school” and is “associated with a significant increase in the child’s earnings in adulthood.”).
 Dahl & Lochner, supra note 218, at 1949.
 The absence of children means that the couple will not reduce its liability through dependency exemptions, thus the marriage bonus is more pronounced.
 One could argue that the very existence of an income tax is a disincentive to work. I do not accept this argument, but I do accept the argument that income splitting currently provides a rational incentive for one spouse to work less or not at all, because the additional income negates the marriage bonus as it increases. Separate filing would place all individuals in similar situations with respect to income and tax liability, because the existence of the additional income would not impact the liability of a spouse.
 I have purposely ignored the earned income credit in this calculation so as to make a comparison only of the benefits of income splitting in this case. If the marriage bonus is calculated as including the earned income credit, then the bonus is $2,156, because the married couple filing jointly would receive an earned income credit of $1,126 while the unmarried filer would receive a credit of only $71. This is due to the fact that the income phase out for the credit is higher for married taxpayers. However, if both couples earned $50,000, or if both couples earned $40,000 but had no children, then neither couple would be eligible for the earned income credit.
 Zelenak, supra note 24, at 398–99. He concludes his article by noting that “[t]here is no absolutely right or wrong way to tax married couples . . . . Whatever the merits of joint returns may have been for mid-twentieth century America, the joint-return system fits poorly with American attitudes and living patterns at the close of the century.”). Id. at 404–05.
 Pew Research Ctr., The Decline of Marriage and Rise of New Families 38 (2010). The report notes that the divorce rate rose sharply after 1960, peaked in 1979, and has declined since that time. The report further suggests that the decline in the divorce rate reflects the simultaneous decline in marriage. Id.
 Each child would allow the husband a dependency exemption of $3,700 in 2011. Depending on the husband’s income, he may also be eligible for a child tax credit of up to $1,000 in 2011. The wife earning a few thousand dollars, on the other hand, would have no tax liability and would not earn enough to receive any portion of the child tax credit. Without qualifying children to claim, her earned income credit could not exceed $464 and might be less; it would roughly offset her FICA taxes, which was the original legislative intent of the earned income credit.
 Zelenak, supra note 24, at 400–01.
 See George K. Yin & Jonathan Barry Forman, Redesigning the Earned Income Tax Credit Program to Provide More Effective Assistance for the Working Poor, 59 Tax Notes 951 (1993).
 Id. at 957–60.
 IRS Restructuring: Hearings on H.R. 2676 Before the S. Comm. on Fin., 105th Cong. 142 (1998).
 Many Code sections provide special benefits for married couples filing jointly and would need to be amended accordingly. See, e.g., I.R.C. § 121(b)(2) (2012) (providing special rules for joint filers who wish to claim the exclusion on gain from the sale of a principal residence).
 Pew Research Ctr., supra note 224, at 11 (reporting a gap of sixteen percentage points in 2008 in marriage rates between college graduates (64%) and those with a high school diploma or less (48%)).
 See supra notes 217–19.