TCHARITY REALLY DOES BEGIN AT HOME: FLORIDA TAX REVIEW Volume 12 2012 Number 8 629 AN EMPIRICAL STUDY OF INNOCENT SPOUSE RELIEF: DO COURTS IMPLEMENT CONGRESS’S LEGISLATIVE INTENT? by Stephanie Hunter McMahon∗ ABSTRACT Under existing law spouses are jointly and severally liable for taxes assessed with respect to their joint income tax returns. As a result, the IRS may pursue either spouse for any taxes owed on those returns. Because Congress was concerned that the IRS was seeking taxes from the “wrong” spouse under the joint and several liability regime, it expanded relief for “innocent” spouses in 1998. Many critics of this relief complain that, as it is applied, the statute offers too little relief to spouses, generally wives, who sign returns while being deceived or compelled by their mates. However, there has been no empirical study of whether the current relief is, in fact, what Congress intended. This article fills the void by first evaluating the provision’s legislative history to determine what relief Congress intended to provide when it acted in 1998. The article then examines the 444 cases appealing for relief under this provision in order to evaluate whether judges are deciding cases invoking the provision consistent with that congressional objective. This article’s empirical study of the success and failure of the innocent spouse provision from Congress’s perspective concludes that the courts are generally applying innocent spouse relief as Congress intended. I. INTRODUCTION ............................................................................. 630 II. THE LAW IN DEVELOPMENT ....................................................... 635 A. Background ............................................................................ 636 B. Congressional Action ............................................................. 639 III. THE LAW IN PLAY ....................................................................... 645 ∗ Associate Professor, University of Cincinnati College of Law. The author would like to acknowledge the comments on previous drafts of this article from the 2011 Law and Society Conference, the 2011 Critical Tax Conference, the 2011 Junior Tax Conference, and the University of Cincinnati faculty workshop, in particular Chris Bryant, Paul Caron, Kristin Kalsem, Francine Lippman, Susie Morse, and Michael Solimine, as well as comments from Dorothy Brown, Margaret Drew, and Carl Smith and for the financial support of the Harold C. Schott Foundation. 630 Florid Tax Review [Vol.12:8 A. Mechanics of the Study ............................................................ 647 1. Choice of Courts ........................................................ 648 2. Breakdown of Provisions ............................................ 654 3. Summary .................................................................... 660 B. Characteristics of Requesting Spouse .................................... 661 1. Sex ............................................................................. 661 2. Marital Status ............................................................ 663 3. Other Characteristics ................................................ 665 4. Summary .................................................................... 668 C. Nature of the Tax Burden ....................................................... 669 1. Unfair Burden ............................................................ 669 a. Source of Liability ........................................ 669 b. Knowledge of Liability ................................. 673 c. Benefit from Liability .................................... 679 2. Crushing Burden ....................................................... 685 a. Economic Hardship ...................................... 685 b. Not Borne by Others ..................................... 689 3. Summary .................................................................... 692 D. Characteristics of Non-Requesting Spouse ............................ 693 1. Abusive ...................................................................... 693 2. Legally Obligated Under State Law ........................... 698 3. Intervening ................................................................. 702 4. Summary .................................................................... 704 IV. CONCLUSION ................................................................................ 705 PART I. INTRODUCTION When Kathleen Sullivan married Joseph Alioto in 1978, she probably thought her future was secure.1 She was the well-educated daughter of the owner of the New England Patriots and he was a prominent antitrust attorney who had spent eight years as mayor of San Francisco.2 The couple “enjoyed a loving, supportive, and harmonious marital relationship, and Mayor Alioto believed it was his absolute duty to care and provide for his family.”3 However, after Joseph’s death in 1998 that duty was shown to have gone unmet. Although when they married Kathleen knew that Joseph was fighting the Internal Revenue Service (IRS), little did Kathleen know when they filed their joint tax returns each year that Joseph was not paying their federal 1. Alioto v. Commissioner, 96 T.C.M. (CCH) 63, 64, T.C.M. (RIA) ¶ 2008- 185 at 986. 2. Id., 96 T.C.M. (CCH) n.3, 65 n.7, T.C.M. (RIA) ¶ 2008-185 at 985–87. 3. Id., 96 T.C.M. (CCH) 63, 64, T.C.M. (RIA) ¶ 2008-185 at 986. 2012] Empirical Study of Innocent Spouse Relief 631 taxes.4 Only when handling Joseph’s estate did Kathleen learn of their total tax liability.5 By that time, she had been a homemaker to the wealthy attorney and politician for twenty years. For the next decade Kathleen fought this tax liability that, by the time of her final resolution, totaled $1,985,511.6 The IRS argued that Kathleen was liable for this tax because she had signed the couple’s joint returns. The Internal Revenue Code provides that every time a married couple signs a joint return, each spouse becomes jointly and severally liable for paying tax on the income that is reported, or failed to be reported, on that return.7 This means that if a couple chooses to file jointly, the IRS can collect from either spouse the taxes due. Spouses remain jointly and severally liable even after they divorce or one spouse dies. Congress has created several exceptions to this joint and several liability, and it was an exception in section 6015 of the Code that ultimately provided Kathleen relief.8 As discussed in Part II of this article, Congress intended this relief for wives who were unfairly left oppressive tax burdens by their divorced or deceased husbands. Primary responsibility for granting or denying this “innocent spouse” relief was given to the IRS, but the judiciary was given oversight over administrative denials.9 Kathleen twice appealed to courts before she finally won.10 4. Id. At the time of their marriage Joseph was fighting the IRS over charitable deductions. Alioto v. Commissioner, 40 T.C.M. (CCH) 1147, T.C.M. (P- H) ¶ 80,360. The couple also had “protracted negotiations” regarding their 1978 return. Alioto v. Commissioner, 67 T.C.M. (CCH) 2133, T.C.M. (RIA) ¶ 1994-051 at 94–235. The IRS satisfied the couple’s 1993 and 1994 liabilities from Joseph’s estate. Kathleen had previously been granted innocent spouse relief for 1989, 1990, and 1991 pursuant to a stipulated decision in docket No. 3013-95. Alioto, 96 T.C.M. (CCH) 63, 66, T.C.M. (RIA) ¶ 2008-185 at 990. 5. In Dec. 1996, the government seized over $2 million in community property to satisfy prior tax obligations but Kathleen still thought the family’s net worth was over $16 million. Alioto, 96 T.C.M. (CCH) 63, 65, T.C.M. (RIA) ¶ 208- 185 at 987. 6. Id., 96 T.C.M. (CCH) 63, 64, T.C.M. (RIA) ¶ 2008-185 at 986. 7. I.R.C. § 6013(d)(3). 8. I.R.C. § 6015; Alioto, 96 T.C.M. (CCH) 63, 71, T.C.M. (RIA) ¶ 2008- 185 at 997. 9. I.R.C. § 6015(e). The IRS created factors to consider when applying the provision, and courts often incorporate these factors in their opinions. Rev. Proc. 2003-61, 2003-2 C.B. at 296, superseding Rev. Proc. 2000-15, 2000-1 C.B. 447. The IRS has recently updated the factors it will consider when applying section 6015(f). Notice 2012-8, 2012-4 I.R.B. 309. This change does not affect the findings of the study, although the revised factors should cause the IRS not to contest relief under fact patterns in which they would have contested relief under Revenue Procedure 2003-61.Carl Smith argues that courts should not rely on the Treasury Department’s factors. Carlton M. Smith, Innocent Spouse: Let’s Bury That “Inequitable” Revenue Procedure, 131 TAX NOTES 1165 (2011) [hereinafter Smith, “Inequitable” Revenue 632 Florid Tax Review [Vol.12:8 That Kathleen was widowed and left with a staggering tax bill that her husband’s estate could not satisfy worked in favor of her claim for relief. Moreover, the court concluded, “During the years in issue Mrs. Alioto reasonably believed that Mayor Alioto was a man of wealth, a man who was on top of everything, and a man in control.”11 Because she reasonably expected Joseph to pay the taxes, the court placed all of the blame for the unpaid tax on her husband. Congress last liberalized the innocent spouse rules granting Kathleen relief in 1998.12 That year, 95 percent of married couples filed jointly, approximately 49 million couples of whom 1.25 million were assessed additional taxes.13 For some of those 1.25 million couples, the system is thought to have failed because the liability for taxes owed was imposed on the “wrong” spouse.14 This can happen for many reasons: The “right” spouse is hard to locate, no longer has money to pay the tax, or has funds that are harder to collect. In 1998, the Senate’s sponsor of section 6015 estimated that 50,000 women were held jointly and severally liable for their husbands, and the Governor’s Accounting Office estimated that 35,000 spouses were held liable who had separated or divorced from the person with whom they had filed.15 It was to help these spouses that Congress first enacted, and then liberalized, innocent spouse relief.16 Since 1998, section 6015 contains three means to relief: a limited equitable relief for those meeting statutory requirements; an allocation of liability for divorced, widowed, or separated spouses; and a general equitable relief to be granted by the Secretary of the Treasury. A review of cases invoking this innocent spouse relief provision contributes to existing scholarship in two significant ways. First, it provides an empirical study of one of the ten issues most litigated by the IRS. This is valuable not only because of the issue itself but, more generally, empirical Procedure]. The issue should not be whether the courts use the factors but how they do so. To the extent courts give substance to the factors, they develop the law for the IRS to apply. 10. Alioto, 96 T.C.M. (CCH) 63, 64, T.C.M. (RIA) ¶ 2008-185 at 986. 11. Id., 96 T.CM. (CCH) 63, 70, T.C.M. (RIA) ¶ 2008-185 at 995. 12. Internal Revenue Service Restructuring and Reform Act of 1998, Pub. L. No. 105-206, § 3201(a), 112 Stat. 685, 734-40 (1998). 13. TREASURY DEPARTMENT, REPORT TO THE CONGRESS ON JOINT LIABILITY AND INNOCENT SPOUSE ISSUES (1998), 9 [hereinafter REPORT ON JOINT LIABILITY]. 14. Id. at 21. 15. 144 CONG. REC. 4474 (1998); U.S. GENERAL ACCOUNTING OFFICE, GAO/T-GGD-98-72, ALTERNATIVES FOR IMPROVING INNOCENT SPOUSE RELIEF [hereinafter GAO, ALTERNATIVES]. 16. See supra Part II. 2012] Empirical Study of Innocent Spouse Relief 633 research in taxation remains an under-developed area, despite an increased focus on this type of research in recent years.17 Unlike prior empirical research focused on statutory interpretation, judicial motivations, or taxpayer responses, this study furthers the scholarly agenda by examining whether, and to what extent, courts implement congressional intent as described in a statutory provision’s legislative history.18 This new focus offers a unique ability to examine the operations of, and the interaction between, the branches of the federal government. For this purpose, Part II draws congressional intent from the Congressional Record and committee reports. Although there are risks with assuming these sources contain Congress’s intent with respect to the innocent spouse provision, the consistency of views expressed therein suggests that there was a dominant vision of what this provision was 17. NATIONAL TAXPAYER ADVOCATE, 2010 ANNUAL REPORT TO CONGRESS 414 (2011) [hereinafter NTA, 2010 ANNUAL REPORT]. See also Michael J. Bommarito II et al., An Empirical Survey of the Populations of U.S. Tax Court Written Decisions, 30 VA. TAX REV. 523 (2011) [hereinafter Bommarito, Empirical Survey]; Assaf Likhovski, The Duke and the Lady: Helvering v. Gregory and the History of Tax Avoidance Adjudication, 25 CARDOZO L. REV. 953, 971–72 (2004); Daniel M. Schneider, Empirical Research on Judicial Reasoning: Statutory Interpretation in Federal Tax Cases, 31 N.M. L. REV. 325, 325 (2001); Leandra Lederman, Which Cases Go To Trial?: An Empirical Study of Predictors of Failure to Settle, 49 CASE W. L. REV. 315 (1999); Michael A. Livingston, Reinventing Tax Scholarship: Lawyers, Economists, and the Role of the Legal Academy, 83 CORNELL L. REV. 365, 368 (1998); Nancy Staudt, Empirical Taxation, 13 WASH. U. J.L. & POL’Y 1, 2 n.8, lists all empirical tax articles from 1993 to 2002. 18. For normative support of the use of legislative intent, see Thomas W. Merrill, The Common Law Powers of Federal Courts, 52 U. CHI. L. REV. 1, 32–33 (1985); Martin Redish & Theodore Chung, Democratic Theory and the Legislative Process: Mourning the Death of Originalism in Statutory Interpretation, 68 Tul. L. Rev. 803 (1994); Daniel Rodriguez & Barry Weingast, The Positive Political Theory of Legislative History: New Perspectives on the 1964 Civil Rights Act and its Interpretation, 151 U. PA. L. REV. 1417 (2003). For a discussion of judges’ decision- making in tax cases, see Bommarito, supra note 17; Nancy Staudt et al., Judging Statutes: Interpretive Regimes, 38 LOY. L.A. L. REV. 1909, 1911 (2004); Schneider, supra note 17; Lederman, supra note 17; John Coverdale, Text as Limit: A Plea for a Decent Respect for the Tax Code, 71 TUL. L. REV. 1501 (1997); Michael A. Livingston, Practical Reason, “Purposivism” and the Interpretation of Tax Statutes, 51 TAX L. REV. 677 (1996); Deborah Geier, Interpreting Tax Legislation, 2 FLA. TAX REV. 492 (1995) [hereinafter Geier, Interpreting]; Michael A. Livingston, Congress, the Courts, and the Code: Legislative History and the Interpretation of Tax Statutes, 69 TEX. L. REV. 819 (1991); Lawrence Zelenack, Thinking about Nonliteral Interpretations of the Internal Revenue Code, 64 N.C. L. REV. 623 (1986). 634 Florid Tax Review [Vol.12:8 intended to accomplish.19 The article then analyzes whether courts effectively use that intent to apply the facts and circumstances tests laid out in the statute. Second, this article contributes specifically to the current public policy debate on innocent spouse relief, most of the debate being critical of joint and several liability.20 One critic concludes that joint and several liability is “generally inequitable on its face” and others contend that innocent spouse relief is a “failure” or has become a “guessing game,” and in the midst of this debate the Treasury Department liberalized its interpretation of one of the three section 6015 tests in January 2012.21 While some of what 19. See Lawrence Solan, Private Language, Public Laws, 93 GEO. L.J. 427 (2007); Daniel Rodriguez & Barry Weingast, The Paradox of Expansionary Statutory Interpretation, 101 NW. U. L. REV. 1207 (2007); William Buzbee, The One-Congress Fiction in Statutory Interpretation, 149 U. PA. L. REV. 171 (2000); Kenneth Shepsle, Congress is a “They,” Not an It, 12 INT’L REV. L. & ECON. 239 (1992). For a discussion of congressional intent in the tax area, see Steven Dean & Lawrence Solan, Tax Shelters and the Code, 26 VA. TAX REV. 879, 903 (2007); Deborah Grier, Interpreting, supra note 18, at 503–04; Michael Livingston, Congress, Courts and the Code, 69 TEX. L. REV. 819 (1991). 20. For articles in law reviews since the 1998 change see J. Abraham Gutting, Note, The “Price” is Right: An Overview of Innocent Spouse Relief and the Critical Need for a Uniform Approach to Interpreting Knowledge Requirement of Internal Revenue Code § 6015, 2 CHARLESTON L. REV. 751 (2008); M. Megan Kerns, Note, Duress: A Perplexing Barrier to Relief from Joint and Several Liability, 58 HASTINGS L. J. 1123 (2007) [hereinafter Kerns, Duress]; Richard Beck, Failure of Innocent Spouse Reform, 51 N.Y.L. SCH. L. REV. 928, 932 (2006) [hereinafter Beck, Failure]; Adrianne Hodgkins, Comment, Getting a Second Chance: The Need for Tax Court Jurisdiction Over IRS Denials of Relief under Section 66, 65 LA. L. REV. 1167 (2005); Lily Kahng, Innocent Spouses: A Critique of the New Tax Laws Governing Joint and Several Tax Liability, 49 VILL. L. REV. 261 (2004); Svetlana G. Attestatova, Note, 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 (2003) [hereinafter Attestatova, Bonds of Joint Tax Liability]; Kari Smoker, Comment, IRS Restructuring and Reform Act of 1998: Expanded Relief for Innocent Spouses—at What Cost? A Feminist Perspective, 60 OHIO ST. L. J. 2045 (1999); Amy C. Christian, Joint and Several Liability and the Joint Return, 66 U. CIN. L. REV. 535, 535 (1998) [hereinafter Christian, Joint and Several Liability]. In addition, one practitioner recently completed a how-to book for seeking innocent spouse relief. ROBERT B. NADLER, A PRACTITIONER’S GUIDE TO INNOCENT SPOUSE RELIEF: PROVEN STRATEGIES FOR WINNING SECTION 6015 TAX CASES (2011) [hereinafter NADLER, INNOCENT SPOUSE RELIEF]. 21. Christian, Joint and Several Liability, supra note 20, at 536; Beck, Failure, supra note 20, at 931; Steve Johnson, SHOULD CONGRESS REFORM THE 1998 REFORM ACT: THE 1998 ACT AND THE RESOURCES LINK BETWEEN TAX COMPLIANCE AND TAX SIMPLIFICATION, 51 KAN. L. REV. 1013, 1058 (2003). 2012] Empirical Study of Innocent Spouse Relief 635 is written reviews a subset of the cases analyzed below, none makes a systematic evaluation of them. Consequently, the authors’ normative assessments often assume an empirical result. Part III of this article provides empirical data for those engaged in this debate and an examination of whether courts are implementing the law as Congress intended. A content analysis of the 444 cases on innocent spouse relief decided between July 22, 1998, the effective date of the latest round of legislative change, and April 15, 2011, requires an examination of these cases’ murky facts and circumstances. This analysis finds that, although there remains uncertainty as to how a particular case will be decided ex ante, courts are doing a relatively good job implementing Congress’s intent. However, courts have not developed consistent interpretations of the factors the Treasury Department uses to define that intent. Part IV concludes with reasons for these results; it does not make a normative evaluation of whether this relief is sufficient or whether courts should implement a statute in accordance with its legislative history. The normative evaluation will come in the second part of a two-part project. This first part examines whether the cases handed down since 1998 show that the regime is accomplishing its legislative purpose, and the second part will assess whether we should be satisfied with that result or whether we should prefer one of the proffered alternatives. Thus, this article is an objective analysis of the law as it operates and judges the success and failure of the innocent spouse law from Congress’s perspective. PART II. THE LAW IN DEVELOPMENT The Internal Revenue Code requires anyone liable for federal income tax to file a federal tax return.22 Married couples are allowed to calculate their liability by filing jointly.23 If couples choose to file jointly, both spouses are required to sign the joint return; however, failure to sign the joint return is not an absolute bar to its validity.24 Spouses’ intent is the dispositive factor.25 Couples may choose to file jointly for many reasons. Joint filing generally offers favorable tax brackets (compared to filing as married filing separately), the ability to claim certain tax credits, administrative convenience, and other real or perceived advantages. 22. I.R.C. § 6011(a). 23. I.R.C. § 6013(a). If spouses do not file jointly, they are required to file as married persons filing separately. I.R.C. § 1(d). 24. Reg. § 1.6013-1(a)(2). 25. See Malkin v. United States, 3 F.Supp. 2d 493 (D.N.J. 1998); Crew v. Commissioner, 44 T.C. Memo. 1145 (1982); Estate of Campbell v. Commissioner, 56 T.C. 1 (1971). 636 Florid Tax Review [Vol.12:8 A separate question from how married couples file their returns is how does the IRS collect the liability due? Each spouse is currently jointly and severally liable for the joint return.26 From the collection perspective, once a joint return is filed, the resulting taxes are not “his” or “her” taxes but “their” taxes, even if only one spouse earns the income reported on the return and even if only one spouse participates in the return’s preparation. A. Background The Treasury Department has consistently supported joint and several liability, but only congressional action made this result certain. The Treasury Department explained its original imposition of joint and several liability in 1923 on the grounds that “a single joint return is one return of a taxable unit and not two returns of two units on one sheet of paper.”27 However, in the 1935 case of Cole v. Commissioner,28 the Ninth Circuit refused to accept the executive branch’s conclusion, arguing that the joint return did not cause spouses to lose their individual identities for tax purposes.29 The language of Cole left open the possibility that Congress could impose joint and several liability, and Congress did so in 1938.30 The House concluded, “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.”31 After having watched years of litigation, Congress gave the executive branch a reprieve from future litigation on the subject.32 In the first five decades of the income tax, the only means for overcoming joint and several liability was for a spouse to prove that he or she signed the return under duress.33 This defense is still available but 26. I.R.C. § 6013(d)(3); Reg. § 1.6013-4(b). 27. T.D. 1882, 15 Treas. Dec. Int. Rev. 203 (1913). 28. Cole v. Commissioner, 81 F.2d 485 (9th Cir. 1935). 29. Id. 30. Revenue Act of 1938, Pub. L. No. 75-554, § 51(b), 52 Stat. 447, 476 (1938); H.R. REP. NO. 75-1860, at 29–30, 48 (1938); H.R. CONF. REP. NO. 75-2330 (1938). 31. H.R. REP. NO. 75-1860, at 30. 32. Thereafter the Court treated the joint return as creating a single unit liable for the collective taxes. Taft v. Helvering, 311 U.S. 195, 198 (1940); Helvering v. Janney, 311 U.S. 189, 192 (1940). 33. Reg. § 1.6013-4(d). “Duress” in section 6015(c)(3)(C) is interpreted as abuse and not legal duress, although that is not how Senator Bob Graham, author of the provision, used the phrase in the Congressional Record. Reg. §1.6015-3(c)(2)(v); 144 CONG. REC. S4473 (1998). With a finding of duress, there is no joint return on which to impose joint and several liability. See Gormeley v. Commissioner, 98 2012] Empirical Study of Innocent Spouse Relief 637 remains hard to prove because it is a subjective analysis.34 In addition, courts have held that the victim spouse must prove not just abuse but that the joint tax return was signed under duress.35 Although duress does not have to be as extreme as receiving a threat of death immediately prior to signing a return, there must be a constraint of will so strong that it makes a person reasonably unable to resist a demand to sign. In the early 1970s, Congress decided that the duress defense was insufficient after several cases were decided in which wives were held liable for taxes on funds their husbands had embezzled. These wives were almost always divorced. Of the ten cases handed down between 1965 and 1971 in which the husband was an embezzler, seven of the couples were divorced and one wife was widowed.36 Although these cases did not win significant attention in the popular press, their judges repeatedly called for congressional reform.37 In fact, the Tax Court once lamented, “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 T.C.M. (CCH) 420, 421, T.C.M. (RIA) ¶ 2009-252 at 1859. Raymond v. Commissioner, 119 T.C. 191, 197 (2002). For those successfully claiming duress, spouses are treated as married filing separately, possibly losing credits and becoming subject to higher tax brackets, whereas those claiming innocent spouse relief may be relieved of all liability. M. Meghan Kerns, Duress, supra note 20, at 1144. 34. See In re Hickley, 256 B.R. 814, 825 (2000); Malkin v. United States, 3 F.Supp. 2d 493, 499 (D.N.J. 1998). 35. Hickley, 256 B.R. at 828; Wiksell v. Commissioner, 67 T.C.M. (CCH) 2360, 2368–69, T.C.M. (RIA) ¶ 1994-099 at 94–486 – 87; see also Stanley v. Commissioner, 45 T.C. 555, 562 (1966) (“Proof that a starving man was ordered at gunpoint to eat a piece of bread would not, standing alone, be satisfactory proof that it had been eaten involuntarily.”). 36. Wissing v. Commissioner, 54 T.C. 1428, 1428 (1970); Abrams v. Commissioner, 53 T.C. 230, 231 (1969); Huelsman v. Commissioner, 416 F.2d 477, 478 (6th Cir. 1969); Sharwell v. Commissioner 419 F.2d 1057, 1058 (6th Cir. 1969); Scudder v. Commissioner, 48 T.C. 36, 38 (1967); Davenport v. Commissioner, 48 T.C. 921, 922 (1967); Wenker v. Commissioner, 25 T.C.M. (CCH) 1237, 1237, T.C.M. (P-H) ¶ 66,240 at 1387; Hackney v. Commissioner 24 T.C.M. (CCH) 655, 655–56, T.C.M. (P-H) ¶ 65,127 at 717–18. Of the six cases where the couple remained married, in four the wife was the embezzler. Hauser v. Commissioner, 29 T.C.M. (CCH) 909, 909, T.C.M. (P-H) ¶ 70,207 at 997; Peters v. Commissioner, 51 T.C. 226, 228 (1968); Pridgen v. Commissioner, 26 T.C.M. (CCH) 131, 131, T.C.M. (P-H) ¶ 67,023 at 143; Horn v. Commissioner, 387 F.2d 621, 622 (5th Cir. 1967); Kenny v. Commissioner, 25 T.C.M. (CCH) 913, 913, T.C.M. (P-H) ¶ 66,174 at 1026; Bonner v. Commissioner, 25 T.C.M. (CCH) 517, 517, T.C.M. (P-H) ¶ 66,096 at 580. 37. For the one article in the New York Times, see Elizabeth Fowler, New Tax Rules Aid Innocent Spouse in Case of Fraud on a Joint Return, N.Y. TIMES, Jul. 29, 1971. 638 Florid Tax Review [Vol.12:8 amelioration. It would seem that only remedial legislation can soften the impact of the rule of strict individual liability.”38 When couples remained married, the Tax Court was less sympathetic.39 Not all courts felt impotent to provide at least certain wives relief. The Sixth Circuit complained, “We are not convinced . . . that the statute is so inflexible that an innocent wife who has been victimized by a dishonest husband must be subjected to an additional appallingly harsh penalty by the United States Government.”40 Shortly before Congress acted, the Sixth Circuit began crafting a balancing test to be used when determining whether a wife could be granted relief.41 As these cases progressed through the courts, Congress completed a year of major revisions to the tax code triggered by revelations that 154 wealthy taxpayers had not paid any income tax, and congressional attention continued to focus on improving tax administration.42 Joint and several liability was one part of that administration. After a round of hearings, Congress claimed that the rule of joint and several liability resulted in a “grave injustice” in the administration of the income tax.43 In 1971, Congress legislated relief, which the Treasury Department did not oppose, and thereby averted the need for the Sixth Circuit’s exercise of judicial power.44 This congressional relief was intended as a hardship relief provision for those taxpayers in serious financial difficulty and was never intended to apply to all joint filers.45 The 1971 provision, enacted as section 6013(e) of 38. Scudder, 48 T.C. at 41. 39. “Accordingly, even though Mary may not have known of or benefited from the embezzlement activity (a fact which we tend to doubt), we must nevertheless reject the petitioners’ contention that Mary is not liable for the deficiencies asserted by the respondent.” Hauser, 29 T,C.M. (CCH) 909, 914, T.C.M. (P-H) ¶ 70,207 at 1002. 40. Huelsman, 416 F.2d at 480–81. 41. Sharwell, 419 F.2d at 1061. 42. Comm. Ways and Means and Comm. on Fin., 91st Cong., 1st Sess. Tax Reform Studies and Proposals U.S. Treasury Department, pt. 1, 89–94 (Comm. Print 1969). 43. S. Rep. No. 91-1537, at 2 (1970); H. Rep. No. 91-1734, at 2 (1970). 44. Act of Jan. 12, 1971, Pub. L. No. 91-679, § 1, 84 Stat. 2063, 2063 (1971); S. Rep. No. 91-1537 (1970). Congress had previously enacted at least one private tax bill helping individual couples. See Richard Beck, The Innocent Spouse Problem, 43 VAND. L. REV. 317, 349–50 (1990) [hereinafter Beck, Innocent Spouse Problem]. 45. See S. Rep. No. 91-1537 at 3; H.R. Rep. No. 91-1734 at 3; STAFF OF THE JOINT COMMITTEE ON TAXATION, PRESENT LAW AND BACKGROUND RELATING TO TAX TREATMENT OF “INNOCENT SPOUSES” (JCX-6-98) (1998) [hereinafter JOINT 2012] Empirical Study of Innocent Spouse Relief 639 the Code, only offered relief in cases involving income omitted from the return where the spouse seeking relief could prove that he or she met certain strict requirements.46 Although section 6013(e) was somewhat liberalized in 1984, innocent spouse relief continued to operate as a hardship provision.47 The requesting spouse had to prove that the joint return contained a “substantial understatement” of tax attributable to “grossly erroneous” items of the other spouse; in signing the return, the spouse seeking relief did not know, and had no reason to know, of the understatement; and, under the circumstances, it would be inequitable to hold the spouse seeking relief liable for the understatement. The substantial omission requirement meant that for most taxpayers the omission had to exceed 25 percent of the gross income shown on the return.48 That the requesting spouse did not know or have reason to know was “rooted in the common law of restitution” as a means of ensuring the requesting spouse was “wholly innocent.”49 And a floor amount of tax liability at $500 prevented small claims from gaining relief.50 B. Congressional Action Concerned about the equity of joint and several liability for joint return filers, in 1995, the American Bar Association (ABA) resolved that it be repealed.51 Congress responded to the ABA by directing the General Accountability Office (GAO) and the Treasury Department to study section 6013(e) and to evaluate the ABA’s proposal.52 The inquiry was no longer COMMITTEE ON TAXATION, PRESENT LAW AND BACKGROUND]; GAO, ALTERNATIVES, supra note 15, at 7. 46. I.R.C. § 6013(e)(3) (repealed). 47. Deficit Reduction Act of 1984, Pub. L. No. 98-369, § 424, 98 Stat. 494, 801. For a description of the 1984 changes, see Lisa Edison-Smith, “If You Love Me, You’ll Sign My Tax Returns” Spousal Joint and Several Liability for Federal Income Taxes and the “Innocent Spouse” Exception, 18 HAMLINE L. REV. 102 (1994). 48. I.R.C. § 6013(e)(4)(A) (repealed). 49. Report on Joint Liability, supra note 13, at 16; S. REP. NO. 91-1537, at 6089. 50. I.R.C. § 6013(e)(3) (repealed). 51. ABA, Proceedings of the 1995 Midyear Meeting of the House of the Delegates, 120 No. 1 ANN. REP. A.B.A. 5–6 (1995); Section Resolutions, 13 A.B.A. SEC. TAX’N NEWSL. 13 (1994). See also Domestic Relations Comm., Am. Bar Assoc. Section on Tax’n, Comments on Liability of Divorced Spouses for Tax Deficiencies on Previously Filed Joint Returns, 50 TAX LAW. 395 (1997). On the other hand, the American Institute of Certified Public Accountants proposed a uniform 10% threshold of gross income for all requests. AICPA Proposes Legislative Changes on Tax Treatment of Marriage and Divorce, 67 TAX NOTES 87 (1995). 52. Taxpayer Bill of Rights II, Pub. L. No. 104-168, §401, 110 Stat. 1452, 1459 (1996). 640 Florid Tax Review [Vol.12:8 whether innocent spouse relief worked as hardship relief but whether it provided “meaningful relief in all cases where such relief is appropriate.”53 These departments both concluded that, although existing relief was not perfect, it was best not to limit a spouse’s liability to his or her “share” of the couple’s taxes.54 With this information, the House Subcommittee on Oversight held a day of hearings on innocent spouse relief.55 These hearings were part of broader hearings focused on a complete restructuring of the IRS with the aim of curbing perceived overzealousness in revenue collection. The Republican Congress sought to require the IRS to give renewed attention to taxpayers as “customers,” which meant expanding many programs, late in the process extended to include greater innocent spouse relief.56 By the time the final bill was before Congress, it was thought to give “David the taxpayer an arsenal of powerful slingshots to use against Goliath the IRS.”57 The four witnesses on innocent spouse relief at the Senate Finance Committee Hearings were divorced women.58 These women told stories certain to elicit sympathy for the wives and anger at the tax system: tales of ex-husbands who not only stuck their ex-wives with extraordinary tax burdens but also shirked their responsibility for child support and of an IRS that told one wife that there was no reason to go after her former husband for 53. H.R. Rep. No. 104-506, sec. 401 (emphasis added). 54. U.S. GOV’T ACCOUNTING OFFICE, GAO/GGD-97-34, INFORMATION ON THE JOINT AND SEVERAL LIABILITY STANDARD (1997). Although its recommendations were limited, the GAO noted that the IRS did not receive many requests and “denied most of them.” Id. at 4. The Treasury Department worried that proposals for proportionate liability “would impose increased burdens on taxpayers and the IRS yet would still require some kind of equitable relief provisions in certain egregious situations.” Report on Joint Liability, supra note 13, at 2. The Treasury Department made no estimation of the cost of repeal although it noted such problems as a lack of computing capacity and the need to hire seasonal workers. Id. at 3, 27– 29. 55. Treasury Department Report on Innocent Spouse Relief: Hearing Before the Subcomm. on Oversight of the H. Comm. on Ways and Means, 105th Cong. 1 (1998) [hereinafter Oversight Subcommittee]. 56. H. R. CONF. REP. NO. 105-599, at 252–55 (1998). See also 144 CONG. REC. S1780 (1998); 143 CONG. REC. H10027, H10032 (1997). 57. 144 Cong. Rec. H5353 (1998) (statement of Representative William Archer). 58. Senate Committee on Finance, Unofficial Transcript of Finance Hearing on Innocent Spouse Tax Rules, 78 TAX NOTES 1009 (1998) [hereinafter Finance Committee]. There remains a sense that Congress acted because divorced or separated women were frequently targeted for former husbands’ taxes. Michael Schlesinger, Obtaining Innocent Spouse Relief in the Face of the Service’s Propensity to Litigate, 109 J. TAX’N 102, 105 (2008). 2012] Empirical Study of Innocent Spouse Relief 641 taxes owed because it could collect from her.59 For those proposing changes to innocent spouse relief, the concern was that taxpayers who should be receiving relief were unsuccessful in obtaining relief under section 6013(e). Senator Bob Graham complained that section 6013(e) was “theoretical” relief because it was “virtually impossible for the standards of that innocent spouse provision to be met.”60 Senator William Roth declared that “the agency is all too often electing to go after those who would be considered innocent spouses because they are easier to locate, as well as less inclined and able to fight.”61 Thus, in their effort to reform the IRS, some within Congress saw revision of innocent spouse relief as an opportunity to limit the IRS’s ability to collect taxes from some wives.62 “Nine out of 10 innocent spouses are women. Maybe that is because they are more likely to pay up when confronted by the IRS. Maybe it is because women sometimes have fewer resources to defend themselves. In either case, singling out women for abusive collection is just plain wrong.”63 In response to the “horror stories” the Senate had heard, one Senator focused his comments only on divorced or separated wives.64 Others expanded their consideration to the widowed or to wives whose husbands had embezzled from them.65 The one reference to husbands as victims in the Congressional Record was meant to surprise the listeners that this could be a problem for husbands as well.66 For those advocating for a new innocent spouse provision, there was a recognition that liberalized relief would be “fairly expensive . . . in terms of the potential for lost revenue.”67 Nevertheless, there was no discussion on 59. Finance Committee, supra note 58. 60. 144 CONG. REC. S4473-74 (1998) (statement of Senator Bob Graham). 61. Finance Committee, supra note 58. See also 144 CONG. REC. S4028, S4033 (1998). 62. All but one mention of innocent spouse relief in the Congressional Record referred to wives, most often divorced wives. See also 144 CONG. REC. S7647 (1998); 144 CONG. REC. S1072-73 (1998); 144 CONG. REC. S1071 (1998); 144 CONG. REC. S4493 (1998); 144 CONG. REC. S4027 (1998); 144 CONG. REC. S4475 (1998); 144 Cong. Rec. H5354. 63. 144 CONG. REC. S725 (1998) (statement of Senator Jon Kyl). 64. 144 CONG. REC. S7653 (1998) (statement of Senator Jack Murkowski). See also 144 CONG. REC. S7642 (1998). 65. 144 CONG. REC. S7628 (1998) (statement of Senator Judd Gregg); S7634 (statement of Senator Harry Reid). 66. But see 144 Cong. Rec. S4504 (1998) (statement of Senator Olympia Snowe). 67. 144 CONG. REC. S4474 (1998) (statement of Senator Bob Graham). Despite the sense of Congress, there is a general sense in discussions of innocent spouse relief that “the consequences to the fisc are not cause for alarm. . . .” Jonathan 642 Florid Tax Review [Vol.12:8 the floor of the actual cost of this relief. There was also no discussion of cases that might appear less sympathetic on their face — Congress kept referring to wives who were deceived before being left crushing tax burdens, often while caring for the couples’ children.68 In the rush to expand taxpayer protections from the IRS, Congress swept through this change to tax practice without fully vetting the cost or the reach of the new provision. That does not mean that proponents were unaware of, or unconcerned about, potential abuse of innocent spouse relief. There were concerns, and rightly so, that some taxpayers may try to abuse the innocent spouse rules by knowingly signing false returns, or transferring assets for the purpose of avoiding the payment of tax, and then claim to be innocent. Obviously, no one would want to open the door to that type of fraud.69 For all of the cases where there might be abuse of this new provision, proponents of liberalization were quick to claim that “relief will not be available in cases of fraud, or if the IRS proves the taxpayer claiming innocent spouse relief had actual knowledge of an item giving rise to the tax liability.”70 On the heels of this debate and as part of its comprehensive reform of the IRS, Congress repealed section 6013(e) and enacted new section 6015, but the result was not what either house had initially proposed.71 The House of Representatives would have removed the hardship nature of the earlier relief but not otherwise changed the law.72 The Senate, on the other hand, would have allowed all spouses (married and divorced) to apportion liability between them.73 With a relatively free rein because the administration was T. Trexler, Contesting Innocent Spouse Relief: The Intervention Pardox, 126 TAX NOTES 499, 499 (2010) [hereinafter Trexler, Contesting]. 68. See supra notes 59–67. 69. 144 CONG. REC. S4474 (1998) (statement of Senator Alphonse D’Amato). 70. 144 CONG. REC. S7623 (1998) (statement of Senator William Roth). 71. Internal Revenue Service Restructuring and Reform Act of 1998, Pub. L. No. 105-206, 112 Stat. 685, 734 (1998). 72. H. R. REP. NO. 105-364, at 19–20 (1997); JOINT COMMITTEE ON TAXATION, PRESENT LAW AND BACKGROUND, supra note 45. 73. S. REP. NO. 105-174, at 56–57 (1998). It was important to one of the Senate bill’s authors that the bill “would not change the tax tables to eliminate the reduced taxes that many times accompany joint filing;” couples were to enjoy the best of marriage bonuses (for those entitled to them) and separate liability. 144 CONG. REC. S1073 (1998) (statement of Senator Bob Graham). The Senate’s liberalization, assuming no interaction with any other proposal, was estimated to cost 2012] Empirical Study of Innocent Spouse Relief 643 cowed, the last thing the Conference Committee considered was innocent spouse relief and, with the House and Senate versions before it, the Committee reached a compromise providing three distinct means of relief.74 In the Conference Committee’s first new provision, as the House had proposed, the 1971 innocent spouse relief was liberalized into current section 6015(b). Pursuant to section 6015(b), a requesting spouse only has to demonstrate that a tax liability is owed because of an understatement attributable to the other spouse’s erroneous item; that the requesting spouse did not know, or have reason to know, of the existence of that erroneous item; and that, taking into account all the facts and circumstances, it would be inequitable to hold the requesting spouse liable for the taxes due.75 The Conference Committee also incorporated from the Senate’s version of the bill section 6015(c), which apportioned liability for divorced or legally separated spouses or those spouses who have lived apart for the prior twelve months.76 Pursuant to section 6015(c), a requesting spouse’s liability can be limited to his or her share of the couple’s liability. Unlike with section 6015(b), inequity is not a factor under section 6015(c). Instead, section 6015(c) eliminates the presumption of unity for spouses who no longer function as a marital unit.77 The presumption is in favor of this allocation unless the IRS can prove that the requesting spouse had actual knowledge of the tax liability.78 In addition to these two forms of broadened but still limited relief, the Conference Committee added a third form of equitable relief to be granted at the IRS’s discretion.79 Section 6015(f) grants the IRS tremendous $5.157 billion between 1997 and 2007. JOINT COMMITTEE ON TAXATION, COMPARISON OF THE ESTIMATED BUDGET EFFECTS OF H.R. 2676 (JCX-44-98) (1998). 74. Unofficial Transcript of Tax Analysts Program on IRS Restructuring and Reform Act, 2008 TNT 146-50; H. CONF. REP. NO. 105-599, at 251–55. 75. I.R.C. § 6015(b). A spouse can get proportional relief under section 6015(b) if he or she can demonstrate lack of knowledge of the extent of the understatement. I.R.C. § 6015(b)(2). 76. I.R.C. § 6015(c). A widow is treated as though no longer married. H.R. CONF. REP. NO. 105-599, at 252 n16. 77. Bryan Camp, The Unhappy Marriage of Law and Equity in Joint Return Liability, 108 TAX NOTES 1307, 1314 (2005). 78. I.R.C. § 6015(c)(3)(C). Congress had asked the GAO and Treasury Department for an evaluation of binding the IRS to negotiated settlements between divorcing and separating spouses; Congress was not willing to go so far in the final legislation. H. R. REP. NO. 104-506, at 7–8 (1996); sec. 401; Oversight Subcommittee, supra note 55. 79. H. R. CONF. REP. No. 105-599 at 254. 644 Florid Tax Review [Vol.12:8 latitude in providing relief if the prior two provisions are inapplicable.80 This provision allows that “the Secretary may relieve such individual of such liability” if, “taking into account all the facts and circumstances, it is inequitable to hold the individual liable.”81 Section 6015(f) is also the only means to relief for spouses where there is the underpayment of tax on a correct return, as opposed to an understatement of tax on the return.82 This final tripartite relief provision was cobbled together, crafted during a period of tremendous legislative change affecting the business operations of the IRS, and the statutory language lays out no singular vision of when and why particular spouses should be granted relief. There was, however, nothing in the final reports to indicate that there had been a change from the earlier statements made in Congress. Even with the very different approaches proposed by each house, the provision’s goal was to grant relief from liability to divorced or separated wives who were left crushing tax burdens by their nefarious husbands. The 1998 Act required the Treasury Department to quickly draft rules implementing the new provision, as the law was effective upon enactment.83 These rules and regulations then guided the IRS in its determination of whether relief should be granted to particular claims. The Treasury Regulation interpreting the equitable prongs of section 6015(b) and (f) are currently based on factors provided in Revenue Procedure 2003-61, although the IRS has recently proposed changes to the Revenue Procedure for section 6015(f), but under both provisions the IRS considers whether the requesting spouse received a significant benefit, directly or indirectly, beyond normal support from the unpaid taxes; was abused; or will suffer economic hardship if relief is not granted.84 In this 80. I.R.C. § 6015(f). In 2006, the Tax Court was given jurisdiction to review stand-alone equitable relief determinations. Tax Relief and Health Care Act of 2006, Pub. L. No. 109-432, div. C, §408, 120 Stat. 2922, 3061 (2006). 81. I.R.C. § 6015(f). 82. The conference committee would not extend section 6015(c) to reported but unpaid tax but, instead, left that for equitable relief. H. CONF. REP. NO. 105-599 at 254–55 (1998). The IRS treats elections under section 6015(b) and (c) as an application under section 6015(f); however, an application of section 6015(f) relief does not automatically trigger an application of section 6015(b) and (c). Reg. §1.6015-1(a)(2). 83. U.S. GOV’T ACCOUNTING OFFICE, GAO-01-589T, INFORMATION ON SELECTED IRS TAX ENFORCEMENT AND COLLECTION EFFORTS 17 (2001) [hereinafter GAO, INFORMATION]. 84. Rev. Proc. 2003-61, 2003-2 C.B. 296, superseding Rev. Proc. 2000-15, 2000-1 C.B. 447; Notice 2012-8, supra note 9. The Treasury Department’s guidelines for section 6015(f) are more extensive than for section 6015(b) and include certain threshold conditions plus additional levels of factors before the Commissioner will grant a request for equitable relief. Id., at 298. 2012] Empirical Study of Innocent Spouse Relief 645 weighing of factors as directed by Revenue Procedure 2003-61, no single factor is determinative, and the IRS must weigh all of the factors together.85 Nevertheless, under this rule while reason to know of the tax deficiency “will not be weighed more heavily than other factors,” actual knowledge of it is “a strong factor weighing against relief.86 Some courts use these factors as a mathematical equation, adding up those factors that weigh for and those against relief to determine whether relief should be granted.87 Throughout this attempt by the Treasury Department to clarify the statute is an attempt to create rules applying section 6015 to spouses for whom Congress intended relief. The next Part will assess to what extent courts feel the Treasury Department’s rules accomplish that objective.88 In addition, the next Part will evaluate to what extent courts apply congressional intent — the desire to provide relief to divorced, deserted, or widowed wives unfairly left with an overwhelming tax liability created by their former husbands. PART III. THE LAW IN PLAY As the IRS adapted to its new focus on customer service — in the middle of an economic downturn — the IRS received 1,200 applications for innocent spouse relief per week in mid–2000.89 More than 46,000 taxpayers had already made 79,000 applications for relief.90 As the IRS dealt with this onslaught of relief requests, it worked to give substance and meaning to the new provision. From its legislative history, section 6015 was created with the intention of providing relief to “innocent spouses”; however, determining who was “innocent” proved costly to administer.91 The Cincinnati Service 85. Id., at 298. 86. Id. Although the Revenue Procedure authorizes consideration of other factors, little evidence of weighing additional factors is apparent from the cases. Pursuant to Notice 2012-8, actual knowledge will no longer be weighed more heavily than other factors. Notice 2012-8, § 4.03(2)(c)(i), supra note 9. 87. See, e.g., Greer v. Commissioner, 97 T.C.M. (CCH) 1075, 1079, T.C.M. (RIA) ¶ 2009-020 at 112–13. Motsko v. Commissioner, 91 T.C.M. 711, 714-15, T.C.M. (RIA ¶ 2006-017 at 104. Pursuant to Notice 2012-8, no one factor or majority of factors necessarily controls the determination for equitable relief. Notice 2012-8, § 4.03(2), supra note 9. 88. More recent developments in the Treasury Department’s interpretation of this provision are discussed infra. 89. Steve Johnson, The 1998 Act and the Resources Link Between Tax Compliance and Tax Simplification, 51 U. KAN. L. REV. 1013, 1044 (2003) [hereinafter Johnson, The 1998 Act]. 90. Id. 91. Joint Committee on Taxation, Estimated Budget Effects of Titles I-VIII of the Conference Agreement Relating to H.R. 2676, 79 TAX NOTES 1741 (1998). 646 Florid Tax Review [Vol.12:8 Center Integrated Case Processing System (ICP) began operations as the nation’s primary reviewer of relief requests in January 2001.92 Of the 48,461 claims for relief sought in 2005, 42.5 percent were denied without reaching the merits and an additional 29.0 percent were disallowed in full.93 Reviewing these denials, the National Taxpayer Advocate found that 29 percent of the IRS’s rejections were because the liability had already been paid and another 6 percent were rejected because the requesting spouse confused innocent spouse relief and injured spouse relief.94 In 26 percent of cases, a joint return had not been filed or a joint return had been filed but the couple was not married or did not sign the return, and in 19 percent, no return had been filed.95 Many of these errors were the result of taxpayers filing for incorrect years and they likely refiled for the correct years.96 The Treasury Inspector General of Tax Administration found that the IRS properly resolved 94 percent of the cases it reviewed in 2004.97 If ICP denies relief, requesting spouses may appeal to the Appeals Office.98 In fiscal year 2010, Appeals heard 5,341 section 6015 claims (less 92. GAO, INFORMATION, supra note 83, at 18. 93. NATIONAL TAXPAYER ADVOCATE, 2005 ANNUAL REPORT, at 329 (2006) [hereinafter NTA, 2005 ANNUAL REPORT]. This is down from the 48.7% defective on their face in 1999-2001. U.S. GOV’T ACCOUNTING OFFICE, GAO-02- 558, IRS INNOCENT SPOUSE PROGRAM PERFORMANCE IMPROVED; BALANCED PERFORMANCE MEASURES NEEDED 34 (2002) [hereinafter GAO, INNOCENT SPOUSE PROGRAM]. 94. NTA, 2005 ANNUAL REPORT, supra note 93, at 331. Taxes already paid was up from 19.4% in 1999-2001 but confusion with injured spouse relief was down from 6.6%. GAO, IRS’S INNOCENT SPOUSE PROGRAM PERFORMANCE IMPROVED, supra note 93, at 34. 95. NTA, 2005 ANNUAL REPORT, supra note 93, at 331. That the couple did not file a joint return or filed it incorrectly was down from 30.6% in 1999-2001. GAO, INNOCENT SPOUSE PROGRAM, supra note 93, at 34. 96. Not noting the incorrect filings double-counts these taxpayer errors as valid claims. If the problem is an incorrect year, the IRS sends the requesting spouse a letter indicating the years that have joint liabilities and inviting appropriate applications for relief. NTA, 2005 ANNUAL REPORT, supra note 93, at 332 n.32. 97. TIGTA, The Innocent Spouse Review Function, May 2005, Ref. 2005- 40-075, at 5–6. TREAS. INSPECTOR GEN. FOR TAX ADMIN., Ref. No. 2005-40-075, THE INNOCENT SPOUSE CENTRALIZED REVIEW FUNCTION ENSURED ACCURATE RELIEF DETERMINATIONS, BUT IMPROVEMENTS COULD INCREASE CUSTOMER SERVICE 5–6 (2005) [hereinafter TIGTA, ACCURATE RELIEF DETERMINATIONS]. 98. In 2005, each case took on average 192 days to process, 807 days if the application went to Appeals. NTA, 2005 ANNUAL REPORT, supra note 93, at 423. While centralized processing is more efficient, there is concern that its efficiencies cause some cases to be denied the relief Congress intended. Scott Schumacher, Innocent Spouse, Administrative Process: Time for Reforms, 130 TAX NOTES 113 (2011) [hereinafter Schumacher, Administrative Process]. 2012] Empirical Study of Innocent Spouse Relief 647 than 4 percent of all appeals) and 4,610 were closed that year.99 In the Appeals process, additional relief was granted in 35 percent of the claims in 2005.100 In a survey conducted by the ABA Section of Taxation, 73.1 percent of those surveyed believed that the Appeals Office is “generally fair” and, for those handling innocent spouse cases in the two years before the survey, only 17 percent thought the Appeals Office did not exercise independence from the auditors initially denying relief.101 While one commenter found that the Appeals Office “rubberstamp[s] whatever the Service has done,” 62.7 percent were satisfied with the way the Appeals Office handled their innocent spouse cases.102 While 71.5 percent of claims were denied relief, only 2.7 percent of those not deemed ineligible on their face were litigated in the Tax Court.103 Nevertheless, there are complaints that too many cases are not resolved by the administrative process and find their way into the judicial system.104 Although that is a normative assessment not being evaluated in this article, innocent spouse relief was listed as one of the top ten litigated issues in the Taxpayer Advocate’s 2010 Annual Report and for every year since 2001 except for 2003.105 Any requesting spouse may seek relief from the Tax Court as long as a petition is filed no later than ninety days after the IRS mails its final determination notice to the requesting spouse or if the IRS fails to issue a ruling within six months.106 This article examines the 444 litigated cases involving claims for innocent spouse relief that have been decided since the 1998 legislative enactment. It evaluates the standards courts apply against the legislative history. In addition, the article explores which particular factors are most likely to weigh for or against relief. A. Mechanics of the Study This sample includes all recorded federal tax decisions in the LEXIS database handed down between July 22, 1998, and April 15, 2011, that include the words “innocent spouse,” “tax,” and “6015.” Each case was coded for a variety of information and compiled into a spreadsheet, including 99. IRS Data Book 2010, table 21. 100. NTA, 2005 ANNUAL REPORT, supra note 93, at 330. 101. ABA, Section of Taxation, Survey Report on Independence of IRS Appeals, August 11, 2007, 1–2. 102. Id. at 12. 103. NTA, 2005 ANNUAL REPORT, supra note 93, at 329–330 n.23. 104. Schumacher, Administrative Process, supra note 98. Schumacher also complains that denials are often made without explanation. Id. 105. NTA, 2010 ANNUAL REPORT, supra note 17, at 414. 106. I.R.C. § 6015(e)(1). 648 Florid Tax Review [Vol.12:8 information about the parties (education, employment history, role in family finances, etc.), the liability (source, amount, etc.), and the parties’ relation to the liability (knowledge of, benefit from, etc.).107 When cases were appealed, information from earlier or later appeals were used to supplement the spreadsheet. In this content analysis, it is not the goal to rank the importance of variables but to look for patterns within the cases that might otherwise be missed while retaining the language and tone of these varied cases.108 This article is concerned with the results of these opinions and not how judges reached their conclusions or what interpretive technique they adopted. The author acknowledges that there is a limit to the information that can be gleaned from this methodology. First, not all cases that were filed resulted in written opinions. Some were settled and others resulted in unwritten bench opinions.109 Therefore, the opinions used in this sample are not all cases filed or decided in the courts. Second, judicial opinions are available only for those claims for which the taxpayer had the resources and inclination to pursue relief in court. Thus, there is a selection bias in the cases. The litigated cases should be among those claims with the less favorable factors for relief as those with more favorable factors would presumably have been granted relief during administrative review.110 Despite these limitations, this analysis should provide valuable information regarding the implementation of innocent spouse relief. Moreover, this review furthers analysis of the extent to which courts defer to executive agencies and the extent to which they implement congressional intent. 1. Choice of Courts The courts included in this sample are all of the federal courts that review federal tax cases. Federal tax cases are held for trial before the Tax Court, District Courts, and the Court of Federal Claims.111 Decisions of the 107. In the charts below, unless a particular case is referenced, case names are not provided to control for footnote length. This information, as well as the spreadsheet, is available from the author. 108. As a result, this Article does not purport to provide predictive indicators of how a particular case might be resolved. Moreover, the small sample size for many of the issues considered prevent arguing their statistical significance. For a discussion of content analysis, see Mark Hall & Ronald Wright, Systematic Content Analysis of Judicial Opinions, 96 CAL. L. REV. 63 (2008). Although a regression analysis might find additional hidden patterns or associations between factors, because of the sample size and the desire to retain the stories of the requesting spouses themselves, that work is left for a future project. 109. Bommarito, Empirical Survey, supra note 17, at 530. 110. Ideally, this assumption will be tested by a later study at ICP. 111. See 28 U.S.C. § 1345(a)(1); 26 U.S.C. § 7422; Flora v. U.S., 357 U.S. 63 (1958). 2012] Empirical Study of Innocent Spouse Relief 649 Tax Court and District Courts are appealed to the appropriate Circuit Court of Appeals; decisions of the Court of Federal Claims are appealed to the Federal Circuit.112 No innocent spouse case has yet been decided by the Supreme Court. Because courts often considered different factors and issues when a case was on appeal, the 9.5 percent of cases heard on appeal are counted separately from the underlying trial opinion. The final tally of innocent spouse cases was 444.113 The Tax Court issues three different types of decisions: regular, memorandum, and summary. In addition to regular opinions, the Tax Court’s memorandum decisions are not officially published and present nothing more than factual issues.114 Summary opinions are issued when the amount in dispute is $50,000 or less and the taxpayer elects to have the case conducted under a “small tax case” proceeding using simplified rules of evidence, practice, and procedure and waives the right to appeal.115 All Tax Court opinions are included in this sample because they disclose how facts and reasoning interact in the decision-making process. The Tax Court is the only court in which a taxpayer does not have to first pay the liability and then sue for a refund, and about 90 percent of all tax cases are heard in the Tax Court.116 Before 1998, 90 percent of innocent spouse cases were initially heard in the Tax Court; since 1998, 91.8 percent (or slightly more than for all tax cases) have been initiated there.117 112. I.R.C. § 7482; 28 U.S.C. §§ 1294, 1295. 113. There are 492 cases as a result of the search; forty-eight cases are not relevant to this study. Bankruptcy and state court decisions are excluded from the sample. 114. James Maule, Instant Replay, Weak Teams, and Disputed Calls, 66 TENN. L. REV. 351, 368 (1999) [hereinafter Maule, Instant Replay]. 115. I.R.C. § 7463. 116. IRS Data Book 2010, table 27. There are three jurisdictional bases upon which the Tax Court may review a claim for innocent spouse relief. First, section 6015(e) provides that a spouse who has requested relief can petition the IRS’s denial of relief or petition the IRS’s failure to make a timely determination. Such cases are referred to as “stand alone” cases in that they are independent of any deficiency proceeding. Stand-alone basis was extended to section 6015(f) in 2006. See supra note 81. Second, the Tax Court may exercise jurisdiction when a claim is raised as an affirmative defense in a petition for redetermination of a deficiency filed pursuant to section 6213(a). Butler v. Commissioner, 114 T.C. 276, 287–88 (2000). Finally, the Tax Court may exercise jurisdiction when the issue is properly raised in a collection proceeding under sections 6320 and 6330. I.R.C. § 6330(c)(2)(A)(i). 117. Stephen A. Zorn, Innocent Spouses, Reasonable Women and Divorce: The Gap Between and the Internal Revenue Code, 3 MICH. J. GENDER & L. 421, 424–5 (1996) [hereinafter Zorn, Innocent Spouses]. 650 Florid Tax Review [Vol.12:8 CHOICE OF COURTS Tax Court Court of Claims District Court Circuit Court Number of cases 369 2 31 42 Percentage of cases 83.1% 0.5% 7.0% 9.5% Number of wins118 136 1 9 8 Winning percentage 36.9% 50.0% 29.0% 19.0% Not only are most cases initiated in the Tax Court, but 88.3 percent of the total taxpayer wins were in the Tax Court and taxpayers’ Tax Court winning percentage, if you discount the two cases heard in the Court of Claims, is the highest. This is despite the fact that taxpayers do not have to prepay their liabilities in the Tax Court and so weaker cases would presumably be brought there because of the lower cost. Based on the 2010 data for the ten most litigated issues, taxpayers’ overall success rate of 20.6 percent was significantly lower than the 34.7 percent rate in this sample for innocent spouse relief.119 Unlike those studies finding the Tax Court biased in favor of the government, the Tax Court is not dismissive of innocent spouse claims.120 In fact, in United States v. Boscaljon,121 the court ordered that an innocent spouse claim be raised for a 74-year-old wife even though she had not requested such relief.122 However, throwing oneself, or one’s client, on the mercy of the Tax Court does not often work. In Gormeley v. 118. Counted as wins for the requesting spouse and not for an intervenor. 119. NTA, 2010 ANNUAL REPORT, supra note 17, at 417. Because of the IRS’s recent change in position on the statute of limitations for section 6015(f), six cases will result in taxpayer victories. Notice 2011-70, 2011-32 I.R.B. 135. These cases are not counted as winning cases in this sample because the change was beyond the period of review. 120. For the debate regarding the Tax Court’s bias, see Maule, Instant Replay, supra note 114; Deborah A. Geier, The Tax Court, Article III, and the Proposal Advanced by the Federal Courts Study Committee: A Study in Applied Constitutional Theory, 76 CORNELL L. REV. 985, 998–1000 (1991). One study concludes that Tax Court trial judges, but not appellate judges, do exhibit a bias. B. Anthony Billings et al. Are U.S. Tax Court Decisions Subject to the Bias of the Judge? 55 TAX NOTES 1259, 1266 (1992) [hereinafter Billings, Decisions]. 121. 105 A.F.T.R. 2d 1501 (2010). 122. Id. It is unlikely Ms. Boscaljon was granted relief because her house was ultimately foreclosed upon. See 2010 U.S. Dist. LEXIS 46354. 2012] Empirical Study of Innocent Spouse Relief 651 Commissioner,123 the court quoted the requesting spouse’s counsel as stating, “So, what I’m trying to ask the Court here to do is try to help my client out here by finding a way to rule because this is an equitable thing that Congress really wanted to help taxpayers get some ruling from the Court under 6015(e).”124 The court would not overlook that the client had not filed a petition within the 90-day window and, therefore, the court did not have jurisdiction to evaluate the merits of the case.125 The Tax Court, where the majority of innocent spouse cases are heard, is composed of nineteen judges (sixteen judges are currently sitting) who have each been appointed by the President and confirmed by the Senate for fifteen-year terms. Unless the Tax Court sits en banc, decisions by individual judges who hear cases are reviewed by the chief judge and, possibly, the full Tax Court.126 In addition to regular judges, special trial judges may be appointed by the chief judge from time to time. The following chart only includes judges who have ruled on at least five innocent spouse cases. 123. 98 T.C.M. (CCH) 420, T.C.M. (RIA) ¶ 2009-252. 124. Id., 98 T.C.M. (CCH) 420, 421, T.C.M. (RIA) ¶ 2009-252 at 1859. 125. Id., 98 T.C.M. (CCH) 420, 421–22, T.C.M. (RIA) ¶ 2009-252 at 1860. See also Carlton Smith, How Can One Argue ‘It’s Not my Joint Return’ in Tax Court?, 124 TAX NOTES 1266 (2009). The case was appealed to the Third Circuit but, before appeal, the IRS abated assessment because of an improperly filed notice of determination. Email with Carl Smith, Cardozo Tax Clinic, June 5, 2011. 126. Billings, Decisions, supra note 120, at 1266. 652 Florid Tax Review [Vol.12:8 TAX COURT JUDGES HEARING INNOCENT SPOUSE CASES Chiechi Cohen Colvin Dawson Foley Gerber Goeke Appoint- ing president Bush Reagan Reagan Kennedy Clinton Reagan W. Bush Sex Female Female Male Male Male Male Male No. of Cases 14 23 10 9 9 10 19 Taxpayer won 2 4 6 6 3 3 10 Winning percent- age 14.3% 17.4% 60.0% 66.7% 33.3% 30.0% 52.6% Haines Halpern Holmes Jacobs Kroupa Laro Marvel Appoint- ing president W. Bush Bush W. Bush Reagan W. Bush Bush Clinton Sex Male Male Male Male Female Male Female No. of Cases 14 6 12 5 8 8 15 Taxpayer won 7 3 5 3 2 1 7 Winning percent- age 50.0% 50.0% 41.7% 60.0% 25.0% 12.5% 46.7% Nims Ruwe Swift Thornton Vasquez Wells Wherry Appoint- ing president Carter Reagan Reagan Clinton Clinton Reagan W. Bush Sex Male Male Male Male Male Male Male No. of Cases 7 16 9 10 20 10 10 Taxpayer won 2 9 5 2 6 0 1 Winning percent- age 28.6% 56.3% 55.6% 20.0% 30.0% 0.0% 10.0% Armen Carluzzo Couvillion Dean Goldberg Panuthos Appoint- ing president Special trial judge Special trial judge Special trial judge Special trial judge Special trial judge Special trial judge Sex Male Male Male Male Male Male No. of Cases 12 11 17 23 17 16 Taxpayer won 6 4 5 8 6 6 Winning percent- age 50.0% 36.4% 29.4% 34.8% 35.3% 37.5% 2012] Empirical Study of Innocent Spouse Relief 653 From this sample of cases, there is a wide range of success rates before these judges, ranging from 0 percent to 66.7 percent. The political party of the appointing president does not appear to influence the likelihood of a judge’s decisions. The number of female judges is small but it too does not appear to be an indicator, although female judges appointed by Republican presidents have a lower-than-average taxpayer success rate of 17.8 percent. Although taxpayers are more likely to win before Special Trial Judges, at a 37.5 percent rate, the number is not significantly higher than the winning percentage before regular judges, 35.7 percent. Of course, to the extent past decisions are an indicator of future holdings, there are some judges that a taxpayer would rather come before than others. There was a dissent in only eleven cases heard by the full court, or 2.5 percent, despite the lack of agreement as to the meaning of various equitable factors discussed below. All but two of these cases involved primarily procedural issues, namely the scope of judicial review of IRS determinations and whether the regulatorily-imposed two-year limit for appeals under section 6015(f) was permissible. The latter issue has received much attention.127 Although the statutory language of section 6015(b) and (c) expressly imposes a two-year window for taxpayers to apply for relief, until July 2011 the Treasury Department imposed the same limit for claims for relief under section 6015(f).128 The validity of this regulatory limit was repeatedly contested in the courts. In Lantz v. Commissioner,129 a divided Tax Court held the regulation was invalid, a holding that was followed in Manella v. Commissioner,130 but their decisions were overturned by the 127. NTA, 2010 ANNUAL REPORT, supra note 17; Bryan T. Camp, Interpreting Statutory Silence, 128 TAX NOTES 501 (2010); Patrick J. Smith, Gaps in the Seventh Circuit’s Reasoning in Lantz, 128 TAX NOTES 1375 (2010); Robert B. Nadler, Equitable Relief: Time to Level the Playing Field, 133 TAX NOTES 899, 902– 03 (2006); William Brown, Recent Cases Expand Potential for Obtaining Innocent Spouse Relief, 83 PRACTICAL TAX STRATEGIES 86, 90 (2009). 128. Notice 2011-70, supra note 119; Reg. § 1.6015-5(b)(1); T.D. 9003, 2002-32 I.R.B. 294. This was despite a letter from Commissioner Shulman to Jim McDermott, U.S. House of Representatives (Apr. 20, 2011), 2011 TNT 86–34. Moreover, the Treasury Department’s Chief Counsel instructed the IRS attorneys not to seek summary judgment for violating the two-year limit but to continue to argue that relief is unavailable. OFFICE OF CHIEF COUNSEL, C.C.N. CC-2009-012, DESIGNATION FOR LITIGATION: VALIDITY OF TWO-YEAR DEADLINE FOR SECTION 6015(F) CLAIMS UNDER TREAS. REG. § 1.6015-5(B)(1) (2009); OFFICE OF CHIEF COUNSEL, C.C.N. 2010-11, VALIDITY OF THE TWO-YEAR DEADLINE FOR SECTION 6015(F) CLAIMS UNDER TREAS. REG. § 1.6015-5(B)(1) AND THE SEVENTH CIRCUIT REVERSAL OF LANTZ (2010). 129. 132 T.C. 131 (2009), rev’d., 607 F.3d. 479 (7th Cir. 2010). 130. 132 T.C. 196 (2009), rev’d., 631 F.3d 115 (3d Cir. 2011). 654 Florid Tax Review [Vol.12:8 Seventh and Third Circuits.131 Before the IRS removed this limitation, the same issue was pending in four other circuits.132 Not all judges agree as to how various factors for relief should be interpreted, despite most cases not having dissents.133 Nevertheless, because this study tests whether courts implement congressional intent, these judicial decisions are measured against the legislative history. Each decision is taken at face value because the concern is the results of cases rather than the motivations of judges.134 In judging the results, the facts and reasoning given in opinions are assumed to be accurate, understanding that they may not be. 2. Breakdown of Provisions Although congressional intent can be identified in committee reports and the Congressional Record, it is not an operational blueprint. As a result, the IRS and courts disagree as to its application to particular facts. Overruling the IRS, courts granted taxpayers relief, at least in part, in 34.7 percent of the litigated cases. Of the 154 cases in which the taxpayer prevailed, the government did not oppose relief in thirty-six;135 the taxpayer won only in part in twenty-six; and in forty-nine the taxpayer won procedural claims (such as in opposition for summary judgment or for remand). In ninety-one cases, which are 20.5 percent of the total number of section 6015 cases, the requesting spouse won complete relief.136 131. Jones v. Commissioner, 642 F.3d 459 (4th Cir. 2011), also found the regulation valid. Camp, supra note 127, argues that the different purposes of (b), (c), and (f) relief justify Tax Court’s decision in Lantz. Smith, Gaps in the Seventh Circuit Reasoning in Lantz, supra note 127, questions the rationale of the Seventh Circuit. 132. See NTA, 2010 ANNUAL REPORT, supra note 17, at 502. 133. See, e.g., supra Part III.C.1.c. and III.C.2.a. 134. This approach is similar to Bradley W. Joondeph, Exploring the “Myth of Parity” in State Taxation: State Court Decisions Interpreting Public Law, Wash. U. J.L. & POL’Y, 205 (2003). For a discussion of how to judge judicial motivation, see Tonja Jacobi and Matthew Sag, Taking the Measure of Ideology: Empirically Measuring Supreme Court Cases, 98 GEO. L.J. 1, 8 (2009); Carolyn Shapiro, Coding Complexity, Bringing Law to the Empirical Analysis of the Supreme Court, 60 HASTINGS L.J. 477 (2009); Lee Epstein et al., Ideological Drift Among Supreme Court Justices: Who, When and How Important, 101 NW. U.L. REV. 1483 (2007); Daniel Scheider, Empirical Research on Judicial Reasoning: Statutory Interpretation in Federal Tax Cases, 31 N.M. L. REV. 325 (2001). 135. Litigation continues because the non-requesting spouse as intervenor opposes relief for the requesting spouse. See supra Part III.D.3. 136. The categories of winning cases do not equal the total wins because some cases fit in multiple categories. For example, those in which judges did not oppose relief may also count as complete relief. 2012] Empirical Study of Innocent Spouse Relief 655 The means to obtaining relief are through one of three avenues provided in section 6015. Each avenue applies to different, and limited, circumstances. Which of the three provisions offers the best chance of success depends in large part on the facts of the case. As described in the prior Part, sections 6015(b) and (c) offer more automated processes because the statutory and regulatory factors to be considered are significantly fewer than imposed on the broad equitable relief of section 6015(f). This was intentional. The Conference Report for the 1998 Act noted that relief under section 6015(f) was to be an equitable last resort.137 Under section 6015(b), the burden of proof is on the requesting spouse; the taxpayer must prove his or her claim by the preponderance of the evidence.138 Two key characteristics of section 6015(b) cases are that these cases involve only the understatement of taxes owed on a joint return and they require a balancing of factors under its equitable prong. In the eleven successful cases decided on the merits solely under section 6015(b), the court found in each that the requesting spouse had no knowledge of the deficiency. In eight of the nine in which the court mentioned substantial benefit, the court found there was none; and in the four cases in which the court mentioned economic hardship, the court found that it would result for the requesting spouse. It is difficult to decipher other common characteristics of successful claims. Five cases involved couples that were still married, three of which involved investments in a tax shelter. Some requesting spouses were stay-at-home mothers, one was a paralegal, one was a police officer. Like section 6015(b), section 6015(c) involves only the understatement of taxes owed on a joint return; however, there is no balancing of equitable factors. The least subjective means of relief, spouses who request relief under section 6015(c) must be divorced, separated, or widowed, but of the twenty-nine successful cases decided on the merits solely under section 6015(c), two cases involved married couples where the husband was in jail, seemingly in defiance of the regulations that do not treat a temporary absence as a separation.139 With a presumption for the taxpayer, one might expect to see fewer than thirty cases brought by requesting spouses under section 6015(c) alone. Looking at these thirty cases plus the twenty additional cases appealed on multiple grounds that were decided based in part on section 6015(c), requesting spouses won on the merits in thirty-four, plus an additional four on procedural issues. Although the standard for the taxpayer is that he or she must prove the claim by a preponderance of the evidence, the burden of proving the requesting spouse had actual knowledge of the understatement, and thereby not qualified for 137. H.R. CONF. REP. NO. 105-599, at 254 (1998). 138. NADLER, INNOCENT SPOUSE RELIEF, supra note 20, at 83. 139. Reg. § 1.6015-3(b)(3)(i). 656 Florid Tax Review [Vol.12:8 relief, is on the IRS.140 All but three of the successful cases on the merits under section 6015(c) turned on lack of actual knowledge of the deficiency by the requesting spouse.141 Sections 6015(b) and (c) provide relief only for the understatement of tax liability; the tax returns themselves must be incorrect. Section 6015(f) is broader and can provide relief for both the understatement and underpayment of tax. In cases seeking section 6015(f) relief alone, 78.3 percent arose at least in part from the underpayment of tax on correctly filed returns. The burden under section 6015(f) is on the taxpayer to prove the equity of relief.142 However, courts have ruled that the Commissioner’s determination to deny relief under section 6015(f) is subject to de novo review and that the administrative record may be supplemented at trial.143 Not all Tax Court judges agree with this approach and neither does the IRS, which contends that the requesting spouse must show that the Commissioner’s denial of relief was an abuse of discretion.144 As with section 6015(b), section 6015(f) requires a balancing of equitable factors. Of the fifty-five cases granting relief on the merits solely under section 6015(f), in sixteen the court found that the requesting spouse had actual knowledge of the deficiency and, in six, that the requesting spouse had reason to know of it. There was significant division in the cases on whether the requesting spouse would suffer an economic hardship if made liable for the tax. In only four did the court find that the requesting spouse had substantially benefited from the deficiency. Finally, as with section 6015(b), there were no common characteristics of the requesting spouse. 140. NADLER, INNOCENT SPOUSE RELIEF, supra note 20, at 83-84. Under section 6015(c), the burden of proving the appropriate allocation of tax items is on the requesting spouse. I.R.C. § 6015(c)(2). In three of the cases granting section 6015(c) relief, issues of allocation were raised. In two of those cases, the court chastised the IRS for not attempting an allocation. Foy v. Commissioner 89 T.C.M. (CCH) 1299, 1304–05, T.C.M. (RIA) ¶ 2005-116 at 923. Bulger v. Commissioner, 89 T.C.M. (CCH) 1457, 1462–63, T.C.M. (RIA) ¶ 2005-147 at 1140–41. In the final case, the government worked through the allocation provisions when each spouse sought relief, but the court granted only partial relief to the husband. Charlton v. Commissioner, 114 T.C. 333 (2000). 141. Nevertheless, in seven cases the courts noted that the requesting spouse either had constructive knowledge or reason to know of the liability. 142. NADLER, INNOCENT SPOUSE RELIEF, supra note 20, at 83–84. 143. Commissioner v. Neal, 557 F.3d 1262, 1264 (11th Cir. 2010); Ewing v. Commissioner, 122 T.C. 32, 38–39 (2004); Porter v. Commissioner, 132 T.C. 203, 210 (2009), aff’g 130 T.C. 115 (2008). 144. CC-2009-021, 2009 TNT 125–5; CC Notice (35)000-338 (2000). See also Patrick J. Smith, Standards for Tax Court Review in Equitable Innocent Spouse Cases, 2012 TNT 42-6. Toni Robinson & Mary Ferrari, Protecting the Innocent, 2000 TNT 181–100. 2012] Empirical Study of Innocent Spouse Relief 657 Before beginning a more detailed analysis of when relief is granted, it is important to note that not all of the litigated cases are meritorious. Because taxpayers have a right to appeal a denial of relief by the IRS, some claims that are invalid on their face have made their way onto the judicial docket. For example, a threshold question for spouses to win relief from joint and several liability is that spouses must sign or intend to sign joint returns. Although the absence of a joint return is a common reason the IRS initially declines relief, eight cases were decided based on a lack of a joint return and, in each, the court held there was no jurisdiction for granting section 6015 relief. In one, the Ninth Circuit upheld the Tax Court’s denial of jurisdiction for spouses who face joint liability under community property laws but do not file joint returns.145 In a ninth case, one spouse claimed (but failed to win) innocent spouse relief because the other spouse would not agree to file jointly and, as a result, the requesting spouse suffered a larger tax bill.146 Whether a joint return exists is but one question that causes cases to be decided on procedural grounds rather than on the merits. CASES DECIDED ON PROCEDURAL GROUNDS § 6015(b) § 6015(c) § 6015(f) Combination Other Brought under 2 10 47 69 30 Won under 0 2 18 23 7 Of the sample, 158, or 35.4 percent, were resolved based solely on procedural issues, and taxpayers won 31.6 percent of the cases decided on procedural grounds.147 In 50.6 percent of these procedural cases, issues of untimeliness, res judicata, and lack of jurisdiction were the basis of the court’s decision. That so many cases were decided on procedural issues is not necessarily a sign of wasteful litigation. Many of these cases resolved 145. In Christensen v. Commissioner, 523 F.3d 957, 962 (9th Cir. 2008), aff’g 90 T.C.M. (CCH) 642, T.C.M. (RIA) ¶ 2005-299, a husband was jointly liable under community property laws despite filing separately. There is no right to appeal to the Tax Court under section 66 which provides limited community property relief. 146. Rogers v. Commissioner, T.C. Summ. Op. 2010-13. 147. In some cases, claims for some years were resolved on procedural and other years on the merits. The Taxpayer Advocate found that 31% of the cases decided in fiscal year 2010 involved procedural issues, with 55% decided in favor of the government, 36% in favor of the taxpayer, and one split decision; 72% involved an examination of the merits, and, of those, 62% were in favor of the IRS, 27% in favor of the taxpayer, and 12% in split decision. NTA, 2010 ANNUAL REPORT, supra note 17, at 500. 658 Florid Tax Review [Vol.12:8 important questions: for example, they have questioned whether a particular regulatory provision is valid and whether a right to intervene extends to heirs. These developments occur in the courts because of the limited guidance provided by the statute as well as taxpayers’ relatively easy access to the courts. In the discussion below, procedural cases are included in the sample unless otherwise stated. The reason for this inclusion is because, although an exact percentage is unknown, most section 6015 claims are settled.148 Procedural decisions that extend the period for negotiations between requesting spouses and the IRS can operate to the advantage of either party. However, one GAO report found that settlements generally operate to the advantage of requesting spouses.149 Therefore, it is important to include in consideration those decisions that prolong or shorten the process. Nonetheless, more cases, 64.3 percent, were decided on the merits than on procedural grounds, a large number under each subsection. CASES DECIDED ON THE MERITS § 6015(b) § 6015(c) § 6015(f) combination of (b), (c), &/or (f) other Brought under 14 20 110 139 4 Won on any grounds 4 16 40 44 1 Winning percentage 28.6% 80.0% 36.4% 31.7% 25.0% Percentage of total wins 3.8% 15.2% 38.1% 41.9% 1.0% § 6015(b) § 6015(c) § 6015(f) combination of (b), (c), &/or (f) other Won based on section 11 29 54 6 5 Percentage of total wins 10.5% 27.6% 51.4% 5.7% 4.8% Taxpayers used section 6015(f) to seek relief much of the time, relying on the provision, at least in part, in 83.6 percent of their appeals decided on the merits. Taxpayers also tended to appeal under multiple provisions, claiming under multiple provisions 48.4 percent of the time. Of 139 cases brought under a combination of subsections, 108 were either a general section 6015 appeal or an appeal under all three subsections. 148. GAO, INNOCENT SPOUSE PROGRAM, supra note 93, at 30. 149. Of the cases settled between 1996 and 2001, 55% resulted in the taxpayer being absolved of liability, 33% in a reduction of liability, and 12% the liability remained the same. Id. 2012] Empirical Study of Innocent Spouse Relief 659 Judges, on the other hand, tended to provide relief under one subsection and, like taxpayers, judges relied heavily on section 6015(f). Despite some expectation that courts would grant “relatively few” section 6015(f) cases when requesting spouses had been denied relief on other grounds, over 50 percent of the time that relief was granted, it was granted under section 6015(f), which by definition, means the requesting spouse did not qualify under section 6015(b) or (c).150 Courts also used section 6015(c) to grant relief when a taxpayer sought relief under a combination of subsections.151 These results are not static over time. Taxpayers’ reliance on section 6015(f) in cases decided on the merits has increased.152 TAXPAYERS’ RELIANCE OVER TIME The data shows an upward trend in appeals under section 6015(f) after a dip in 2006, but a decline in the combination of appeals after a peak in 2004. A question that cannot be answered from this data is why taxpayers would not always appeal using at least some combination of section 6015(f). 150. Section 6015(f) was relied on alone 55 times and used in conjunction with another subsection 4 more times. Johnson, The 1998 Act, supra note 89, at 1059. See also Smith, “Inequitable” Revenue Procedure, supra note 9. Id. 151. Section 6015(c) relief could be complete if the court allocated all liability to the other spouse. See Reg. § 1.6015-3(d)(4). 152. The chart does not include cases decided in 2011 because the dataset includes a shortened period. 0 5 10 15 20 25 30 §6015(b) §6015(c) §6015(f) 660 Florid Tax Review [Vol.12:8 There are two things to note with respect to taxpayers’ reliance on the three subsections. First, there was a backlog of cases begun before 1998, but this should not greatly affect the conclusions below regarding the implementation of congressional intent because the same analysis should have been applied to these holdover cases. It does mean, however, that there might have been an artificially large number of cases in the early years as taxpayers waited for the more lenient provision before pressing their cases in court. Second, the decrease in the number of section 6015(f) cases in 2006 was likely the result of Ewing v. Commissioner,153 which held the then- existing version of section 6015(e) did not grant the Tax Court jurisdiction to decide subsection (f) claims in stand-alone appeals.154 Congress extended this jurisdiction in December 2006.155 It is harder to draw conclusions based on successful cases because there are few wins in any given year. Nevertheless, some trends can be seen. COURTS’ RELIANCE OVER TIME Much as with taxpayers, judges’ reliance on section 6015(f) has increased over time, although judges were more willing to rely on section 6015(c) than were taxpayers. The number of cases in which judges relied on a combination of subsections has remained small. 3. Summary As with most cases involving federal taxation, the vast majority of innocent spouse cases are litigated in the Tax Court. Although some judges appear to be more likely to grant relief than others, no pattern is discernible regarding which types of judges are more likely to rule in favor of relief. 153. 439 F.3d 1009 (9th Cir. 2006), rev’g 122 T.C. 32 (2004). 154. Id. at 1015. 155. See supra note 79. 0 2 4 6 8 10 12 §6015(b ) §6015(c ) §6015(f) 2012] Empirical Study of Innocent Spouse Relief 661 From the data, courts appear most willing to grant relief under section 6015(c) and, thereby, to apportion liability between spouses. However, courts are not unwilling to use section 6015(f) and to go to the merits of these cases in deciding whether relief is warranted. The factors gathered from the cases will be discussed more fully below. This data allows us to evaluate the extent to which courts implement the legislature’s intent. As seen already, the broadening of the relief under section 6015(f) was intended by Congress; the expectation was that the equitable provision would give taxpayers a final means of relief and taxpayers and judges are willing to use it as such. Below, the article examines whether or not divorced, separated, or widowed wives who were unfairly left crushing tax burdens by their nefarious husbands — women in the factual situation with which Congress was concerned — are more or less likely to be granted relief. B. Characteristics of Requesting Spouse When Congress debated expanding innocent spouse relief in 1998, its focus was on aiding divorced, separated, or widowed wives.156 The question addressed in this section is to what extent do those granted relief fit this characterization. 1. Sex Relief from joint and several liability is often perceived, both in Congress and by academics, as relief for women.157 Before the 1998 change in law, 90 percent of petitioners for innocent spouse relief were women.158 At that time, men won ten of the forty-two cases (or 23.8 percent) that they brought at the trial level and women won ninety-two of 393 cases (or 23.4 percent) that they brought.159 Although their trial level winning percentages were similar, men never won on appeal and women won sixteen times (but had two opinions in their favor overturned).160 In the period since 1998, women have continued to bring most cases for innocent spouse relief. The following chart includes all cases decided on whatever grounds, grouped based on whose behalf relief was claimed. For 156. See supra notes 59–67; Finance Committee, supra note 58; Oversight Subcommittee, supra note 55; 144 CONG. REC. S7647 (July 8, 1997). 157. See supra notes 59–67. 158. Zorn, Innocent Spouses, supra note 117, at 424. 159. Id. at 425 n.10. 160. Id. at 425 n.10. 662 Florid Tax Review [Vol.12:8 example, if a husband claimed relief as beneficiary of his wife’s estate, it was coded as a wife’s suit.161 GENDERED RELIEF Brought Trial Case Won Trial Case Wife Husband Both162 Wife Husband 338 59 4 128 16 Brought Appeal Won Appeal 37 5 0 8 0 From the evidence, Congress was right to identify innocent spouse relief as a women’s issue. Wives sought relief in 85.4 percent of total cases, 85.3 percent of the trial cases and 88.1 percent of appeals. Not only do women bring more cases, courts appear to be more sympathetic to wives than to husbands. Wives won 21.6 percent of their appeals and 37.4 percent of their trials and husbands won 0.0 percent of their appeals and 25.4 percent of their trial cases. As a result of the dominance wives have in bringing suit, wives won 89.5 percent of total taxpayer victories. Although women are more likely to litigate a claim for innocent spouse relief, both spouses can apply for relief and, if they each win, allocate liability between them.163 Because much relief is granted in earlier administrative phases, it is hard to determine from the available data when both spouses sought relief, although both spouses definitely sought relief in four. If only one spouse is granted innocent spouse relief, the other remains liable for the entire debt. In response to this continuing liability, some non- requesting spouses have protested in court. Courts, however, have been unsympathetic to non-requesting spouses who sought to reduce their liability as a result of the other spouse being granted relief. 161. A non-requesting spouse might contest liability of the estate of a requesting spouse if the statute of limitations has run against the non-requesting spouse. See Jonson v. Commissioner, 353 1181, 1182–83 (10th Cir. 2003). United States v. Boscaljon, 105 A.F.T.R. 2d 1501 (2010) was excluded because the government instigated the request. 162. This column reflects cases when each spouse independently sought relief. 163. JOINT COMMITTEE ON TAXATION, OVERVIEW OF PRESENT LAW RELATING TO THE INNOCENT SPOUSE, OFFERS-IN-COMPROMISE, INSTALLMENT AGREEMENT, AND TAXPAYER ADVOCATE PROVISIONS OF THE INTERNAL REVENUE CODE, 3 (JCX-22-01) (2001). 2012] Empirical Study of Innocent Spouse Relief 663 2. Marital Status As much as Congress identified innocent spouse relief as a women’s issue, Congress also identified it with divorced or separated women.164 The requesting spouse’s marital status at the time the couple filed the return and at the time appeal was made to the courts can often be determined. For purposes of the following chart, the couple was coded as separated, divorced, or widowed if a couple was separated, divorced, or widowed for at least one year for which a claim of relief was made. If the court noted that a spouse was in the process of separating or divorcing, the couple was coded as separated or divorced. These two choices highlight the number of requesting spouses who were not in traditional relationships at the time of filing the return or are more consistent with congressional sympathies at the time of trial. MARITAL STATUS OF THOSE SEEKING RELIEF At time of filing At time of trial Wife requesting Husband requesting Wife requesting Husband requesting Married 312 44 99 8 Separated (legally or physically) 31 14 12 5 Divorced 11 3 210 45 Widowed 6 2 44 5 Other/never legally married 20 4 13 4 Excluding those spouses for whom marital status is unknown or who were never married, most couples were married (86.6 percent) when they filed but divorced (59.6 percent) when they took their case to trial.165 Nevertheless, forty-eight spouses (13.3 percent) filed the joint return when widowed, separated, or divorced from the non-requesting spouse. Also a significant number, 107, or 25.0 percent, sought relief from joint and several liability while still married to the non-requesting spouse. Many of these couples, 25.2 percent, faced liabilities as a result of an investment in a tax shelter and 29.9 percent faced liabilities as a result of unpaid taxes. In 86.7 percent of the litigated cases in which couples remained married and knowledge of the unpaid tax or the unreported item was raised, the requesting spouse was deemed to have some amount of knowledge, with 164. See supra text at notes 55–65. 165. The National Taxpayer Advocate reported that of those seeking relief in 2001, 34% were single filers and 51% filed as “head of household.” NTA, 2005 ANNUAL REPORT, supra note 93, at 328. Thus, 85% were unmarried. 664 Florid Tax Review [Vol.12:8 40 percent having actual knowledge of the tax deficiency. These spouses are not consistent with the stereotype portrayed by Congress when it enacted expanded relief in 1998. Examining the marital status of those winning relief, winning at either the trial court or on appeal is coded a victory. If a spouse won on a procedural matter or if a spouse won only in part, the claim was also coded as victorious. MARITAL STATUS OF WINNING TAXPAYERS At time of filing At time of trial Wife requesting Husband requesting Wife requesting Husband requesting Married 116 11 17 1 Separated (legally or physically) 11 5 6 1 Divorced 3 0 92 14 Widowed 1 0 19 0 Other/never legally married 6 2 4 1 As Congress focused heavily on divorced, separated, and widowed spouses in 1998, so too the courts are more sympathetic to those spouses at trial. Excluding those for whom marital status was not one of the traditional categories, 132 of 150, or 88.0 percent, of successful spouses were separated, divorced, or widowed at trial. Only 14.2 percent of those who remained married were successful at trial. At the same time, it is best to have been married when filing the return; 81.9 percent of those who were successful were married at the time of the filing. Some of these opinions reflect a traditional view of marriage. In Korchak v. Commissioner,166 the court questioned: Should she be punished for being a loving, trusting wife, a homemaker and mother…? Had she asked any questions about Madison Recycling, her husband and the accountant would have reassured her. . . . It would be egregious to take away her retirement at an age when she earned that right. The innocent spouse relief was designed for these circumstances.167 166. 92 T.C.M. (CCH) 199, T.C.M. (RIA) ¶ 2006-185. 167. Id., 92 T.C.M. (CCH) 199, 209, T.C.M. (RIA) ¶ 2006-185 at 1272. 2012] Empirical Study of Innocent Spouse Relief 665 But meeting the traditional congressional archetype is not always enough to win relief. In Torres v. Commissioner,168 an immigrant woman was held responsible for her former husband’s debts despite the IRS conceding that she had no knowledge of the understatement of liability.169 Although the court considered the case to be close, that she significantly benefited from the understatement and would not suffer an economic hardship from paying the tax “constrain us to conclude that it would not be inequitable to hold petitioner liable.”170 Based on a balancing of equitable factors, the court held this wife liable.171 Few husbands claiming relief lived in non-traditional arrangements. Nevertheless, husbands seeking relief who relied on their wives to handle family finances won relief 38.5 percent of the time, which is more often than husbands normally won. On the other hand, in Maluda v. Commissioner,172 an estranged wife took the money the husband claimed was designated to pay the tax attributable to her husband’s income.173 The court did not grant him relief.174 In Stewart v. Commissioner,175 that the husband knew his wife was employed was sufficient to overcome his claim that his wife handled all of the family’s finances. 3. Other Characteristics Many members of Congress depicted innocent spouses as “women, most of them working moms struggling to make ends meet.”176 A review of common characteristics of requesting spouses should shed light on whether the courts shared an image of those worthy of relief. From the facts of the cases, there is nothing distinctive about wives or husbands who seek relief. For example, in twenty-six cases, requesting wives were teachers or former teachers, in nineteen they were nurses or former nurses, in one she was a 168. T.C. Summ. Op. 2009-170. 169. Id. 170. Id. 171. Id. 172. 98 T.C.M. (CCH) 545, T.C.M. (RIA) ¶ 2009-281. 173. Id. 174. 98 T.C.M. (CCH) 545, 546, T.C.M. (RIA) ¶ 2009-281 at 2040. The husband stipulated to unfavorable facts. While the court wanted to grant relief, it felt constrained not to. Id. 175. See T.C. Summ. Op. 2010-31. 176. 144 CONG. REC. S4493 (1998) (statement of Senator Spencer Abraham). See also CONG. REC. H10003 (1997) (statement of Representative Jerry Weller), S1072 (statement of Senator Al D’Amato), S4473 (statement of Senator Bob Graham), H5356 (statement of Representative Johnson); Finance Committee, supra note 58 (statement of Chairman Johnson). 666 Florid Tax Review [Vol.12:8 producer at ABC and another two had PhDs, and in five they had attended law school. In two cases the requesting husband was guilty of a crime related to the tax filing; in eight the husband was trained as a lawyer or had (or was obtaining) his MBA; and in four the wife had embezzled the unreported income. Thus, there were many different types of people who requested innocent spouse relief. The following chart provides information regarding the education level of the requesting spouse. EDUCATION LEVEL Requesting Relief Winning Relief Husband Wife Total Husband Wife Total Less than high school 2 9 11 1 7 8 (72.7%) High school / GED 4 39 43 0 19 19 (44.2%) Some college or less than 4 year degree 4 28 32 3 13 16 (50.0%) College degree 12 69 81 3 23 26 (32.1%) Post- graduate 8 15 23 1 5 6 (26.1%) Spouses did better than average when the court mentioned the requesting spouse’s education level. Compared to the total average winning percentage of 34.7 percent, these spouses won 39.5 percent of their claims. The amount of education and the perception of education also matter, with those having lesser education generally doing better. In three of the four cases in which the court found the requesting spouse was well or highly educated, judges denied relief. For the three claims with graduate degrees in business, one won in small part. Of the eight claims for relief made by lawyers, all lost. Because Congress depicted innocent spouses as those who were struggling with the tax burden (sometimes as housewives and sometimes as single working mothers), the following chart examines their employment, both at the filing and at the trial.177 177. Compare Finance Committee, supra note 58, at [8] and [176]. 2012] Empirical Study of Innocent Spouse Relief 667 EMPLOYMENT Requesting Relief Winning Relief Husband Wife Total Husband Wife Total Non-requesting spouse primary earner at filing 2 72 74 0 36 36 Non-requesting spouse not primary earner at filing 38 169 207 11 58 69 Not employed at trial 2 47 49 1 25 26 Employed at trial 33 148 181 9 48 57 From a small sample, when husbands requested relief and their wives were the couples’ primary earners, husbands never won relief. More surprisingly, if husbands requested relief when their wives were not the primary earners, husbands won 28.9 percent of the time. On the other hand, when wives requested relief and their husbands were primary earners, wives won relief 50.0 percent of the time; and if the non-requesting husband was not the primary earner, the wife won 34.3 percent of the time. Judges also granted relief to 53.1 percent of those requesting spouses noted not to be employed at the time of the trial and only to 31.5 percent of those who were then employed. Finally, Congress concluded, “perhaps most egregious of all . . . [collection] efforts are often undertaken without regard to the impact that they will have on the welfare of the innocent children involved. . . .”178 CARE FOR DEPENDENT CHILDREN Requesting Relief Winning Relief Husband Wife Total Husband Wife Total Caring for children 10 73 83 2 35 37 Almost 45 percent of requesting spouses who were noted as caring for dependent children were granted relief. Although not mentioned by Congress, one characteristic that might be important for winning relief in the courts is whether the requesting spouse is represented by counsel at trial.179 178. Id. at [22]. 179. For fiscal year 2010, thirty-six section 6015 cases were appealed to the courts, of which 56% of the requesting parties were pro se. This was the lowest 668 Florid Tax Review [Vol.12:8 REPRESENTATION Requesting Relief Winning Relief Husband Wife Total Husband Wife Total Represented 22 182 204 4 68 72 Pro se 45 198 243 13 69 82 Wives (47.9 percent) are more likely than husbands (32.8 percent) to be represented when appealing the denial of innocent spouse relief. However, representation does not appear to be a critical matter for determining whether a spouse wins. In total, 35.3 percent of represented spouses won; 33.7 percent won when they were pro se. On the other hand, wives won 37.4 percent of the time when they were represented and 34.8 percent when they were pro se; husbands won 18.2 percent when they were represented but 28.9 percent when they were pro se. Included in those who were pro se are eleven cases in which either the requesting spouse or the non-requesting spouse, if the couple remained married, was an attorney. In those eleven cases (of which six wives were the requesting spouse), the requesting spouse lost each time. 4. Summary From the available evidence, courts appear to share Congress’s expectations that wives will request innocent spouse relief and that certain types of wives are more likely to be the intended beneficiaries of relief. For example, having a marital status both when filing the return and litigating in court that conforms to stereotypes can be helpful in winning relief. The existence of traditional marital relationships does not always work for or against requesting wives, although husbands who take untraditional roles find it hard to win relief. Although other characteristics of requesting spouses vary greatly, those spouses most likely to win are those that conform to congressional archetypes, such as those who are unemployed or care for dependent children. Those without a high school education are more likely than their percentage among the ten most litigated issues. NTA, 2010 ANNUAL REPORT, supra note 17, at 416. Of the twenty cases in which the taxpayer was pro se, the taxpayer prevailed in whole or in part six times, or 30%. Id. at 417. When the taxpayer was represented, in sixteen cases, the taxpayer prevailed eight times, or 50%, the most successful among the top ten issues. Id. Leandra Lederman and Warren B. Hrung, Do Attorneys Do Their Clients Justice? An Empirical Study of Lawyers’ Effects on Tax Court Litigation Outcomes, 41 WAKE FOREST L. REV. 1235, 1282 (2006), conclude that represented clients do better in Tax Court litigation. 2012] Empirical Study of Innocent Spouse Relief 669 more educated counterparts to win relief. The need for representation is less clear, although it is more helpful for wives than husbands. C. Nature of the Tax Burden In its 1998 debates, Congress concentrated on spouses burdened by an “unfair obligation” who “have become financially wiped out when they find themselves liable for taxes, interest, and penalties because of actions by their spouse of which they were unaware.”180 This section explores to what extent do those granted relief fit this characterization. 1. Unfair Burden a. Source of Liability As Congress considered the circumstances surrounding the generation of tax liabilities, so too might those circumstances influence whether a court perceives a burden as unfair. For example, courts might be less sympathetic if the deficiency arose from an investment in a tax shelter or if large, or small, amounts of revenue are at stake. On the other hand, that the requesting spouse’s subsequent refunds are used to pay an old debt might be more sympathetic than if the spouse is litigating other tax issues and innocent spouse relief is only one of many taxpayer defenses. The section 6015 cases reflect many sources of liability. In cases with multiple issues, each issue was counted separately. SOURCE OF LIABILITY Unpaid tax Unreported income Disallowed tax shelter Disallowed deductions Other Not provided Number of cases 180 110 50 45 41 40 Percentage of cases 38.6% 23.6% 10.7% 9.7% 8.8% 8.6% Number won 65 36 14 22 12 14 Winning percentage 37.8% 32.7% 28.0% 48.9% 29.3% 35.0% Most cases seeking innocent spouse relief involved properly reported but unpaid tax. With respect to these appeals, courts granted relief more often than the average success rate for innocent spouse cases of 34.4 percent. Tax returns that understated couples’ income are both appealed less frequently than unpaid tax cases and have a lower winning percentage, despite the fact 180. 144 CONG. REC. S1072 (statement of Senator D’Amato); S4511 (1998) (statement of Senator Dodd). 670 Florid Tax Review [Vol.12:8 that underreporting of taxes constitutes a much larger proportion of the gross tax gap than the underpayment of taxes.181 There were significantly fewer cases involving disallowed deductions, but their winning percentage was much higher. Of the twenty-four cases when courts noted the culpability of requesting spouses beyond that of mere knowledge of the tax deficiency, requesting spouses lost all of their claims. On the other hand, in the ten opinions noting that the requesting spouse was not culpable in the tax evasion, the requesting spouse won 60 percent of the time. For investments in tax shelters specifically, spouses who had invested in tax shelters appealed an IRS denial of relief fifty times. Requesting spouses involved with the invalidation of tax shelters were granted relief 28.0 percent of the time. One law firm, Merriam, Pierson, and Gellner, litigated sixteen cases that arose from tax shelters, each involving buyers of interests in fraudulent partnerships, and won one case in which the IRS conceded it did not prove the requesting spouse had actual knowledge of the investment and three cases for litigation costs because the IRS did not initially grant relief. In general, a significant source of tax avoidance is self-employment income. Self-employed taxpayers have relatively low overall compliance rates, at 43 percent, compared to those who earn wages, where taxpayers pay approximately 98 percent of their taxes.182 Because of these lower compliance rates, self-employed spouses might unfairly leave their mates with unpaid tax bills. In 103 cases, or 23.2 percent of all section 6015 cases, the non-requesting spouse was self-employed and courts granted relief in 50.5 percent. On the other hand, in twenty-seven cases, 6.1 percent, the requesting spouse was self-employed and, in that context, won relief in only 22.2 percent. Thus, self-employment by non-requesting spouses appears to increase the chance of relief being granted while the self-employment of requesting spouses decrease it. Who prepares the couple’s tax return may also play a role in how a court will perceive the source of liability. For the following chart, twenty- seven returns were counted as both spouse-prepared and professional / software prepared if one spouse was stated to work closely with the return preparer or to use the software. 181. Underreporting of the individual income tax constitutes $235 billion and the underpayment of all taxes is $46 billion. IRS News Release, IRS estimates $450 Billion Gross Tax Gap for 2006, Table 1, 2012 TNT 5-51. 182. IRS News Release, IRS Updates Tax Gap Estimates, IR 2006-28 (Feb. 14, 2006) (accompanying charts). Nina Olson, Olson Testifies on Fairness in IRS Enforcement, 2007 TNT 44–28. 2012] Empirical Study of Innocent Spouse Relief 671 RETURN PREPARER Wife prepares & seeks relief Wife prepares/ husband seeks relief Husband prepares & seeks relief Husband prepares/ wife seeks relief Profes- sional preparer /software Shelter promoter Number of cases 17 8 13 42 135 17 Percent- age of cases 3.8% 1.8% 2.9% 9.5% 30.4% 3.8% Number won 6 6 3 22 50 5 Winning percent- age 35.3% 75.0% 23.1% 52.4% 37.0% 29.4% In 239, or 53.8 percent, of section 6015 cases, the court did not indicate who prepared the return. However, in thirty cases that do make this observation, the spouse seeking relief also prepared the return. In these cases, the requesting spouse’s success rate was 30 percent, but women did significantly better than men. On the other hand, a wife requesting relief when her husband prepared the return won 52.4 percent of the time, but a husband requesting relief when his wife prepared the return won 75 percent of the time. Spouses’ success rate when a shelter promoter prepared the return is less than spouses’ success rate in cases involving tax shelter investments generally. As shown below, who handled family finances was used as an indicator of knowledge of the source of the liability in 34.7 percent of cases. FAMILY FINANCES Wife controls finances & seeks relief Wife controls finances / husband seeks relief Husband controls finances & seeks relief Husband controls finances / wife seeks relief Separate finances Both controlled Number of cases 39 13 2 73 12 15 Percentage of cases 25.3% 8.4% 1.3% 47.4% 7.8% 9.7% Number won 10 5 0 43 11 2 Winning percentage 25.6% 38.5% 0% 58.9% 91.7% 13.3% Judges noted most frequently when the husband controlled the finances and the wife sought relief, and they granted relief in these cases 58.9 percent of 672 Florid Tax Review [Vol.12:8 the time.183 Similarly, when wives controlled the finances and husbands sought relief, courts granted relief more than the average success rate.184 Courts are most likely to grant relief when judges note that couples keep their finances separate. If both spouses participate in family finances (or husbands handle the finances and then seek relief), relief is much harder to obtain. Similarly, although the subset is small, working in the same business with the non-requesting spouse affects the courts’ decisions, unless the requesting spouse is clearly a dependent worker. For example, in Sykes v. Commissioner,185 the wife kept the records for her husband’s law practice and was held liable for taxes on income the business generated. On the other hand, In Harper v. Commissioner,186 the wife sometimes drove patients at a substance abuse treatment facility operated by her husband but otherwise had no knowledge of the business’s finances, and she was held not responsible for knowing of the operation’s income. Finally, what caused the requesting spouse to initiate the request might impact judges’ perception of the tax burden. In 17.1 percent of section 6015 cases, the court noted that the government had used the requesting spouse’s refund, garnished wages or accounts, or imposed a levy on property in order to satisfy the liability. In 7.6 percent of those cases, the court granted relief. In 12.8 percent of all section 6015 cases, the court noted that the government had issued a notice of determination or intent to collect and, in 33.3 percent of these cases, the court granted relief. Finally, in 4.1 percent of cases the court noted that the requesting spouse was litigating other tax issues and, in 44.4 percent, the court granted relief. This result could lead one to conclude that judges do not weigh this factor significantly when considering relief for requesting spouses. 183. For example, in Doyle v. Commissioner, 94 Fed. Appx. 949 (3d Cir. 2004), despite having only a high school education, the wife was held to know about a horse breeding tax shelter because of her role writing family checks and handling household expenses. “Nancy certainly should have been alerted to the prospect that ‘something is rotten in the state of Denmark’ when she signed a tax return that deducted almost 70% of the couple’s gross income–something that on its face reduced the family’s apparent income to a level totally at odds with the couple’s lifestyle.” Id. at 952. 184. When a husband relied on his wife to handle the family’s finances, as his mother had done when he was a child, the court held he still had a duty to inquire. Molsbee v. Commissioner, 98 T.C.M. (CCH) 331, 332, T.C.M. (RIA) ¶ 2009-231 at 1732. 185. 98 T.C.M. (CCH) 105, 151, 156, T.C.M. (RIA) ¶ 2009-197 at 1473– 74, 1480. 186. T.C. Summ. Op. 2010-153. 2012] Empirical Study of Innocent Spouse Relief 673 b. Knowledge of Liability Committee reports, and many statements made on the congressional floor, stressed that an innocent spouse has no knowledge of his or her spouse’s actions.187 In two of the methods for obtaining relief under § 6015, Congress included the requesting spouse’s knowledge of the tax deficiency as a factor in determining whether the spouse was unfairly saddled with a tax burden.188 Congress did not include knowledge of liability in the more open- ended equitable relief of section 6015(f); but, under the Treasury Department’s rules interpreting section 6015(f), knowledge is included as an equitable factor.189 For cases decided on the merits (so that knowledge would be a consideration), the requesting spouse’s knowledge was raised in 269 cases, or 93.7 percent. Moreover, in six cases decided on procedural grounds, for which knowledge is not a deciding factor, the court noted that the requesting spouse had knowledge or reason to know and, sided with the government in all but one. Therefore, it is unsurprising that when a wife conceded knowledge but argued it was “legally irrelevant” the court disagreed.190 Whether imposed by Congress or the Treasury Department, what the knowledge standard is depends on the type of relief the requesting spouse is claiming. For purposes of section 6015(b) and (f), whether the requesting spouse knew, or had reason to know, of the understatement (or underpayment for section 6015(f)) is determined by whether the IRS can prove the spouse actually knew of the deficiency or whether a reasonable person in similar circumstances would have known.191 The IRS may deny relief if the taxpayer had reason to know of the cause of the deficiency, although it is only one factor to be weighed pursuant to the equity prong. Until the IRS’s recent proposed changes, however, actual knowledge, was “a 187. See H. Rep. No. 105-174; H. Rep. 105-599; 144 Cong. Rec. S4492 (statement of Barbara Boxer), S4500 (statement of Dianne Feinstein), S1072 and H7623 (statement of William Roth), and S7653 (statement of Carol Moseley-Braun). 188. I.R.C. § 6015(b)(1)(C); I.R.C. § 6015(c)(3)(C). Knowledge is tested at the time the original return was filed to encourage married couples to file amended returns. See Billings v. Commissioner, 94 T.C.M. (CCH) 183, 186–87, T.C.M. (RIA) ¶ 2007-234 at 1426 (2007). 189. Rev. Proc. 2003-61, 2003-2 C.B. 296, 298; Notice 2012-8, § 4.03(2)(c), supra note 9. 190. Mellen v. Commissioner, 84 T.C.M. (CCH) 530, 538, T.C.M. (RIA) ¶ 2002-280 at 1703 (2002). 191. Reg. § 1.6015-2(c). Before 1998, the courts gave the explanation that spouses had “reason to know” for denying relief approximately 55% of the time. Zorn, Innocent Spouses, supra note 117, at 425. 674 Florid Tax Review [Vol.12:8 strong factor weighing against equitable relief” in cases where taxes are understated on the return.192 How to judge what one should reasonably know depends both on the cause of the tax liability and how the judge applies the test. For cases of underpayment of taxes on correctly filed returns, the IRS questions whether the requesting spouse had reason to know the non-requesting spouse would not pay the tax.193 Similarly, the IRS considers whether the requesting spouse did not know or have reason to know of the item giving rise to the deficiency when taxes are understated.194 Courts’ applications of these tests are not so clear, particularly when evaluating understatements. For example, for income understated on the return, the question can be whether the requesting spouse knew of the underlying transaction producing the income or should have known (or questioned) about additional income based on family expenses.195 For overstated deductions, that requesting spouses must know of the underlying transaction can be applied as whether a reasonable person looking at the return would question whether the deduction was odd and so trigger an obligation to inquire or as whether the requesting spouse had actual knowledge of the transaction that produced the overstated deduction.196 Unlike section 6015(b) and (f), in section 6015(c), which provides for apportioned relief, Congress placed the burden on the IRS to prove that the requesting spouse had actual knowledge of the understatement to the satisfaction of the courts.197 Actual knowledge is meant to be a higher standard than the reason to know test provided in section 6015(b) and (f); however, scholars complain that it has been interpreted by the courts to require only knowledge of the underlying transaction and that this 192. Rev. Proc. 2003-61, 2003-2 C.B. 296, 298; Notice 2012-8, § 4.03(2)(c)(i), supra note 9. 193. Id. 194. Id. The Regulations adopt a facts and circumstances test. Reg. § 1.6015-2(c), 3(c)(2)(B)(1). 195. Compare Mitchell v. Commissioner, 292 F.3d 800, 803-04 (D.C. Cir. 2002) with Ohrman v. Commissioner, 86 T.C.M. (CCH) 499, 503, T.C.M. (RIA) ¶ 2003-301 at 1666 (2003), aff’d, Ohrman v. Commissioner, 157 Fed. Appx. 997 (9th Cir. 2005). 196. Compare Hopkins v. Commissioner, 121 T.C. 73, 80 (2003) with Phemister v. Commissioner, 98 T.C.M. (CCH) 163, 171, T.C.M. (RIA) ¶ 2009-201 at 1503 (2009). In Price v. Commissioner, referenced in forty-seven cases in this sample, the Ninth Circuit reasoned that since erroneous deductions are necessarily reported on a tax return, any spouse who signs the joint return is put on notice that an income-producing transaction occurred. 887 F.2d 959, 965 (9th Cir. 1989). 197. I.R.C. § 6015(c)(3)(C). This has been interpreted to mean actual knowledge of the factual circumstances which gave rise to the erroneous deductions. See King v. Commissioner, 116 T.C. 198, 204 (2001); Reg. § 1.6015-3(c)(2)(B)(2). 2012] Empirical Study of Innocent Spouse Relief 675 “weaken[s] the intended remedial effect.”198 The Treasury Department states that if the requesting spouse made a “deliberate effort” to avoid learning about the taxes owed or jointly owned the property, the IRS can conclude the spouse had actual knowledge.199 The following chart examines the types of knowledge that courts have found in cases decided on the merits.200 As discussed above, although cases were often brought under a combination of subsections, courts tended to grant relief under only one subsection. FINDINGS OF KNOWLEDGE § 6015(b) § 6015(c) § 6015(f) § 6015(b), (c), and/or (f) Brought Wins Brought Wins Brought Wins Brought Wins Court finds actual knowledge 6 - 3 4 31 13 68 1 Court finds constructive knowledge or reason to know 2 - 2 4 41 6 33 3 Spouse or IRS concedes knowledge 1 - - 0 5 1 2 - Court finds no knowledge 4 11 11 15 26 30 24 5 Unknown; not discussed 1 - 4 3 4 3 5 - When taxpayers won under section 6015(b), the court always found that the requesting spouse did not know or have reason to know of the deficiency. This was not the case for wins under section 6015(f) or section 6015(c), the latter being unexpected because one of section 6015(c)’s requirements is that the requesting spouse must not have actual knowledge of the deficiency. For 56.6 percent of requesting spouses who won under section 6015(f), the court found no knowledge, but in 37.7 percent of the cases won under section 6015(f), the requesting spouse either had actual knowledge or reason to know of the deficiency. The results are similar under section 6015(c). For 57.7 percent of requesting spouses who won under section 6015(c), the court found no knowledge of the deficiency. However, 198. I.R.C. § 6015(c)(3)(C); Beck, Failure, supra note 20, at 948. See also Attestatova, Bonds of Joint Tax Liability, supra note 20, at 863–64, Senators Bob Graham, Al D’Amato, Dianne Feinstein, and Tim Johnson introduced an amendment to make the actual knowledge test apply at the time the individual signed the return. 144 CONG. REC. S4544 (1998) (Proposed Amendments to H.R. 2676). 199. Reg. § 1.6015-3(c)(iv). Requesting spouses can overcome actual knowledge if they were victims of domestic violence and did not challenge the filing for fear of retaliation. Reg. § 1.6015-3(c)(v). 200. Cases for litigation costs or on other grounds are omitted from the chart. 676 Florid Tax Review [Vol.12:8 in 15.4 percent of the cases in which a spouse won under section 6015(c), the court found actual knowledge; and in 11.5 percent, the court did not discuss knowledge despite it being a requirement for deciding relief on the merits. Courts found constructive knowledge when a requesting spouse did not review the completed return or signed a blank return. Requesting spouses were expected to know what they signed unless they were prevented from reviewing the return. In twenty-six of the cases decided on the merits, judges found that there was constructive knowledge.201 In addition, in Jones v. Commissioner,202 the court found that if a requesting spouse was on notice that the other spouse had unreported income but did not know the exact amount of income, the requesting spouse must fulfill a duty of inquiry or risk being charged with constructive knowledge of the understatement. Nevertheless, the requesting spouse in Jones was granted relief under section 6015(f) after balancing all of the factors weighing for equitable relief. In several cases, spouses were required to exercise greater diligence than simply requesting information. In Cheshire v. Commissioner203 and Wiksell v. Commissioner,204 wives noticed either an ineligible deduction or unreported income on the return and asked about the tax consequences of the mistakes. For that reason, they were held to have actual knowledge of the deficiency and, ultimately, denied relief. Much as ignorance of what is on the return provides little relief, ignorance of the law is unlikely to win relief if it means relieving one spouse completely, despite the ABA’s 1995 proposal that would have allowed ignorance of the law as a defense.205 In the seven cases in which a spouse claimed reliance on bad advice, either from a return preparer or tax software, the resulting ignorance of the law was no excuse for lack of knowledge. In one case, a wife who prepared the couple’s tax return failed to win relief because she had incorrectly relied on an accountant’s opinion that certain of her former husband’s disability income was excluded from tax.206 In another case, the wife signed relying relied on her divorce attorney’s incorrect 201. But see Sunleaf v. Commissioner, 97 T.C.M. (CCH) 1283, 1286, T.C.M. (RIA) ¶ 2009-052 at 413-14 (2009) (Wife who relied on her husband to file the returns and, therefore, did not review them, was granted relief). 202. 99 T.C.M. (CCH) 1457, 1460, T.C.M. (RIA) ¶ 2010-112 at 667 (2010). See also Kruse v. Commissioner, 100 T.C.M. (CCH) 524, 525, T.C.M. (RIA) ¶ 2010-270 at 1636 (2010); Pierce v. Commissioner, T.C. Summ. Op. 2003- 126. 203. 282 F.3d 326, 300 (5th Cir. 2002). 204. Wiksell v. Commissioner, 67 T.C.M. (CCH) 2360, 2367, T.C.M. (RIA) ¶ 94,099 at 94–485 (1994), rev’d, Wiksell v. Commissioner, 90 F.3d 1459 (9th Cir. 1996). 205. ABA, Proceedings of the 1995 Midyear Meeting, supra note 51, at 5– 6. 206. Jaske v. Commissioner, T.C. Summ. Op. 2010-85. 2012] Empirical Study of Innocent Spouse Relief 677 explanation of the creation of loss “credits.”207 In both cases, reliance was to their detriment. Judges also hold that lack of knowledge of joint and several liability is no excuse but with greater reservation. In Kelly v. Commissioner,208 a wife claimed that she believed she had to file jointly because she was married. “On the basis of the level of petitioner’s education, the amount of her control over the filing of the tax returns, and her extensive communication with the tax preparer, the Court finds this argument unpersuasive. Even if petitioner did not have a comprehensive understanding of tax laws, she had reason to know of her joint and several liability for the taxes shown on the joint returns.”209 In Washington v. Commissioner,210 a wife “was under the impression that she was required to file a joint return because she was married at the time” (although she had previously filed as married filing separately).211 After balancing equitable factors, the court granted relief but not on the basis of ignorance of the law. Judges are more sensitive to claims of ignorance of the law when relief is being granted under section 6015(c). In Mora v. Commissioner,212 both spouses relied on a tax shelter promoter but the liability was allocated to the husband. Similarly, in King v. Commissioner,213 the husband, a used car salesman, also owned a cattle ranch. The wife kept the records for the venture and prepared the couple’s tax returns. The Tax Court sustained the wife’s separation of liability under section 6015(c) because the tax treatment of the venture depended upon the husband’s subjective intent to make a profit. Although the wife knew of the business, the IRS was required to prove that the wife knew that the ranch operated as a hobby and not for profit. The Treasury Department supports a stronger position against ignorance of the law defenses for fear that it would necessarily lead to further expansion of this defense: “There is no apparent reason why the tax liability of both spouses should not be excused if they both did not understand the tax consequences of their transaction.”214 In other words, both spouses may be equally ignorant so that it might inequitable to excuse one but not both from the liability. In this sample, no court has responded to this concern. 207. Estate of Gurr, T.C. Summ. Op. 2002-7. 208. 100 T.C.M. (CCH) 507, 512, T.C.M. (RIA) ¶ 2010-267 at 1615 (2010). 209. Id. 210. 120 T.C. 137, 202 (2003). 211. Id. 212. 117 T.C. 279, 281 (2001). 213. 116 T.C. 198, 205-06 (2001). 214. REPORT ON JOINT LIABILITY, supra note 13, at 51. See also T.D. 9003, supra note 128, at 29. 678 Florid Tax Review [Vol.12:8 Ignorance of the law is, therefore, sometimes an excuse. Ignorance of the facts underlying the return tends to be more leniently considered, although this standard is not applied consistently. For example, in Braden v. Commissioner,215 the court held that a husband had no knowledge of the deficiency despite knowing his former wife had received a distribution from her father’s estate. “Although Ms. Braden certainly was in a position to know that the distributions came from her father’s IRA’s, the record does not contain any evidence that she or anyone else told petitioner that the distributions consisted of IRA withdrawals and interest income or gave him any reason to conclude the distributions were taxable.”216 On the other hand, in Charlton v. Commissioner,217 a husband was not allowed to rely upon summaries of Schedule C expenditures provided by his wife when he prepared the couple’s returns. The question runs through these cases of what a spouse, particularly a wife, can be expected to know about family finances. However, the more a requesting spouse is told about family finances, the more likely the spouse is to be held knowledgeable of tax deficiencies.218 It is hard to win relief if a spouse discusses finances with the requesting spouse even if the requesting spouse ultimately defers.219 It is also hard to win relief if the judge finds that the spouse should have known more about the family’s finances. In Alt v. Commissioner,220 a doctor funneled his earnings through dozens of corporations to reduce the couple’s tax liability. The court was unpersuaded by the argument that the doctor’s wife, a college-educated, 74-year-old woman “was reared in a culture that demanded women refrain from questioning [the] breadwinner regarding fiscal matters.”221 Of course, not all couples in traditional relationships that limit access to information fare so 215. 81 T.C.M. (CCH) 1380, 1383, T.C.M. (RIA) ¶ 2001-069 at 527 (2001). 216. Id. 217. 114 T.C. 333, 336 (2000). See also Capehart v. Commissioner, 204 Fed. Appx. 618, 620 (9th Cir. 2006), aff’g, Capehart v. Commissioner, T.C.M. (RIA) ¶ 2004-268 (2004); Grossman v. Commissioner, 182 F.3d 275, 231 (4th Cir. 1999). 218. Compare Sowards v. Commissioner, 85 T.C.M. (CCH) 1517, 1529, T.C.M. (RIA) ¶ 2003-180 at 970–71 (2003), with Pierce v. Commissioner, 85 T.C.M. (CCH) 1553, 1560, T.C.M. (RIA) ¶ 2003-188 at 1007–08 (2003). 219. See Golden v. Commissioner, 548 F.3d 487 (6th Cir. 2008). 220. 101 Fed. Appx. 34, 42 (6th Cir. 2004), cert denied, 543 U.S. 1000 (2004). In one pre-1998 case, Judge Mary Ann Cohen refused to allow a wife to seek innocent spouse relief after her husband pursued a different tax case covering the same tax year. The court found she “implicitly authorized” her husband to pursue the case and assumed that “it would be taken care of” by her husband. Levin v. Commissioner, 71 T.C.M. (CCH) 2938, 2942, T.C.M. (RIA) ¶ 96,211 at 96–1558 (1996). 221. Alt, 101 Fed. Appx. at 42. 2012] Empirical Study of Innocent Spouse Relief 679 badly. In thirty-two cases in which there was a traditional relationship, defined as the requesting spouse not working outside the home (unless working for the non-requesting spouse) and not handling family finances, the requesting spouse won 68.8 percent of the time. From this review, the application of the knowledge or reason to know standard does have a significant amount of inconsistency. Reliance on judges to determine the credibility of witnesses is a necessary part of evaluating this factor. Witness credibility was specifically raised in 110 cases, or 24.8 percent. CREDIBILITY Number of Cases Husband Wins Wife Wins Both credible 4 0 3 Neither credible 7 0 1 Husband credible 13 5 4 Husband not credible 5 0 1 Wife credible 54 0 37 Wife not credible 26 0 4 IRS credible 1 0 0 To the extent that one thinks the judicial system is good at evaluating witnesses, this is less a concern than for those who are more skeptical of judicial wisdom. One thing is clear, the knowledge factor is not a rote check- the-box formulation. c. Benefit from Liability Richard Beck argues that the Treasury Department did not support the repeal of joint and several liability because “it saw hobgoblins of abuse waiting to pounce, in the form of fraudulent schemes where one spouse would transfer all the couple’s assets to the other to put them beyond the I.R.S.’s reach.”222 Congress shared this fear. When Congress expanded innocent spouse relief, it was concerned that taxpayers would transfer assets to avoid paying taxes and that some requesting spouses would unfairly receive tax relief.223 Only those who were unfairly left with a tax bill should be relieved of liability.224 222. Beck, Failure, supra note 20, at 948 n.87. 223. See Finance Committee, supra note 58, at 183 (statement of Sen. John Chafee); 144 CONG. REC. 7623 (1998) (statement of Sen. William Roth); H.R. REP. NO. 105-599, at 253–54 (1998); H.R. REP. NO. 105-364, pt. 1, at 152–53 (1997). 224. See notes 58–68. 680 Florid Tax Review [Vol.12:8 According to the Treasury Department, if a requesting spouse received a significant benefit from the underpayment of taxes, requiring the requesting spouse to pay the tax is not inequitable.225 However, a significant benefit is only one factor to be considered under the equitable prong of sections 6015(b) and (f) as a means of measuring whether the requesting spouse would unfairly benefit from liability relief. There is no reference to significant benefit in section 6015(c), which does not have an equity component, although an allocation is prohibited if spouses transferred property to avoid tax.226 One critic claims that the “law-abiding spouse often receives no financial benefit from the other spouse’s underpayment,”227 but the courts do not share this view. Whether a spouse significantly benefits from a tax deficiency has been held the most important factor of equity.228 In 169 of the cases decided on the merits, the issue of significant benefit was raised. SIGNIFICANT BENEFIT Court finds significant benefit Court finds no benefit IRS concedes no benefit Court finds factor neutral Number of cases 76 68 20 5 Taxpayer wins 35 7 13 0 When the court found that the requesting spouse enjoyed no significant benefit from the tax deficiency, the requesting spouse won relief in 51.5 percent of cases; when the IRS conceded there was no benefit, the requesting spouse won 65.0 percent of the time. On the other hand, when the court found the requesting spouse did benefit, relief was granted in 9.2 percent of cases. The issue of significant benefit was raised more often under section 6015(f) than under section 6015(b) because there are more section 6015(f) cases, but there was no meaningful difference in how the factor was used if the case was won under section 6015(b) or (f). This factor was raised in only two cases decided on section 6015(c) alone, despite section 6015(c) not requiring an equitable analysis. Regulations define whether a requesting spouse significantly benefited from a tax deficiency as whether the spouse received a benefit 225. Reg. § 1.6015-2(d); Rev. Proc. 2003-61, 2003-2 C.B. 296, 299. 226. I.R.C. § 6015(c)(4). 227. Christian, Joint and Several Liability, supra note 20, at 570–71. 228. Cheshire v. Commissioner, 282 F.3d 326, 338 (5th Cir. 2002). 2012] Empirical Study of Innocent Spouse Relief 681 beyond normal support.229 However, the regulations fail to define normal support. Evidence of a benefit may, but does not have to, consist of transfers of property between spouses made at any time.230 The example provided in the regulations is of a spouse receiving life insurance proceeds beyond normal support and the insurance premiums are traceable to items omitted from gross income.231 In the two cases involving life insurance, one found there was a significant benefit and the other did not. In Bozick v. Commissioner,232 the court held that a requesting spouse received insurance proceeds, not because of the tax avoidance, but “because her husband paid for the life insurance policy throughout his lifetime. . . .”233 On the other hand, in George v. Commissioner,234 the court held that if a husband had paid his taxes, there would have been less in the IRA or insurance for the wife to receive on his death. Struggling to apply this factor, judges have not agreed on what it means to create a significant benefit.235 Improving cash flows, even if the money was reinvested in the tax shelter generating the cash flows, is sometimes a significant benefit to both spouses.236 Similarly, paying a child’s college tuition may be a significant benefit to both spouses.237 Thus, as with the knowledge standard, there is uncertainty in the application of this factor. 229. Reg. § 1.6015-2(d). 230. Id. 231. Id. 232. 99 T.C.M. (CCH) 1242, 1244, T.C.M. (RIA) ¶ 2010-61 at 355 (2010). 233. Id. 234. T.C.M. (RIA) ¶ 2004-261 at 1581 (2004). 235. The government expects that the test for determining whether there is an economic benefit to be harder to apply in states where couples live under a community property regime; however, this arose in only one case and, in it, the court held that the wife was unlikely to receive any benefit from community funds. IRS Releases Publication on Innocent Spouse Relief, 2008 TNT 71-63 (April 8, 2008); Haltom v. Commissioner, T.C.M. (RIA) ¶ 2005-209 at 1597 (2005). 236. Compare Capehart v. Commissioner, T.C.M. (RIA) ¶ 2004-268 at 1642 (2004), aff’d, Capehart v. Commissioner, 204 Fed. Appx. 618 (9th Cir. 2006), Juell v. Commissioner, 94 T.C.M. (CCH) 143, 148, T.C.M. (RIA) ¶ 2007-219 at 1372–73 (2007), and Abelein v. Commissioner, T.C.M. (RIA) ¶ 2004-274 at 1718 (2004). See also Casula v. Commissioner, T.C. Summ. Op. 2008-49; Wizen v. Commissioner, T.C. Summ. Op. 2007-99. 237. In the one case where the requesting spouse was being sent to college during the years of tax deficiency, this expenditure was held not to be a significant benefit. Griffin v. Commissioner, T.C. Summ. Op. 2005-41. 682 Florid Tax Review [Vol.12:8 Ambiguity also exists in the weight to be given to this factor and who bears the burden of proof. In Schultz v. Commissioner,238 the court held that when the IRS did not prove that the requesting spouse significantly benefited from the tax deficiency, whether the requesting spouse received a significant benefit was a neutral factor that weighed in favor of relief. Thus, the burden of proof was on the IRS. On the other hand, in Smolen v. Commissioner,239 the court concluded that there was no evidence that the taxpayer did not receive a significant benefit and, therefore, the factor weighed against relief. In Smolen, the taxpayer bore the burden of proof. Both cases were brought under section 6015(f). Perception of the couple’s lifestyle might influence a judge’s willingness to find a significant benefit. When a couple enjoys a lavish lifestyle, it is harder for the requesting spouse to claim there was no such benefit. In twenty-two of the cases in which a significant benefit was found, the court portrayed the couple as well off, and in only one of those cases was the requesting spouse granted relief. The court frequently questions whether there was an extravagant lifestyle and, if not, finds there was no significant benefit regardless of the amount of taxes avoided.240 Whether a judge will find that a couple lived lavishly is often evaluated in comparison to prior years. In Butler v. Commissioner,241 the court held that “although the record demonstrates that petitioner enjoyed a high standard of living during 1992 and maintained accounts at various upscale department stores where she made significant purchases, there is no evidence in the record indicating whether such expenditures were out of the ordinary when compared to petitioners’ spending habits in prior years.”242 On the other hand, unusually lavish expenditures and trips abroad are evidence of receipt of a significant benefit.243 If the requesting spouse receives assets in order to preserve them from creditors, this is often sufficient evidence that the requesting spouse would not be unfairly burdened by the tax liability. For example, in Ohrman 238. 100 T.C.M. (CCH) 353, 355, T.C.M. (RIA) ¶ 2010-233 at 1387 (2010). 239. T.C. Summ. Op. 2010-106. 240. See, e.g., Sjodin v. Commissioner, T.C.M. (RIA) ¶ 2004-205 at 1248– 49 (2004), vacated and remanded, Sjodin v. Commissioner, 2006-1 U.S. Tax Cas. (CCH) ¶ 50,357, 97 A.F.T.R.2d (RIA) 2006-2622 (8th Cir. 2006) 241. 114 T.C. 276, 285 (2000), abrogated, Porter v. Commissioner, 132 T.C. 203 (2009). See also Stolkin v. Commissioner, 96 T.C.M. (CCH) 143, 143 T.C.M. (RIA) ¶ 2008-211 at 1099–1100 (2008) 242. Id. 243. See, e.g., Alt, 101 Fed. Appx. at 43; Doyle v. Commissioner, 94 Fed. Appx. 949, 953 (3d Cir. 2004). 2012] Empirical Study of Innocent Spouse Relief 683 v. Commissioner,244 a couple was legally separated but remained living together. The husband had transferred more than $782,000 in property to his wife one week after the IRS sent the couple a letter stating their tax deficiency. Under the separation agreement, the husband retained only his personal belongings. The court concluded: [P]etitioner’s use of State family law as a vehicle to lend legitimacy to Mr. Ohrman’s transfer of assets and income to her is the type of abuse that Congress expressly intended to stop by adding paragraph (4) to section 6015(c). While the State of Oregon’s equitable distribution rules provided the mechanism for the transfer of Mr. Ohrman’s assets and income to petitioner, they do not negate the principal purpose for which the transfer occurred, the avoidance of tax.245 On this basis, the court denied the wife innocent spouse relief so that the innocent spouse rules would not protect family property. When the requesting spouse was not the transferee, not all transfers failed to protect assets from the IRS. In Wiener v. Commissioner,246 two or three months after the IRS assessed the joint tax liability, the husband transferred the family home to a trust established by his father. The court held: Although we understand respondent’s concern about the timing of the transfer, we reject respondent’s implied argument that the transfer was a transfer between spouses as part of a fraudulent scheme by such spouses within the meaning of Rev. Proc. 2000-15, sec. 4.01(5). The transfer was not between petitioner and Mr. Wiener; it was between petitioner and the Charles Wiener Trust.247 The wife was therefore granted innocent spouse relief, even though she and her husband continued to reside in the home. On the other hand, in Andrews 244. 86 T.C.M. (CCH) at 504, T.C.M. (RIA) at 1666 (2003). See also Doyle v. Commissioner, 85 T.C.M. (CCH) 1108, 1111–12, T.C.M. (RIA) ¶ 2003-96 at 468–69 (2003), aff’d, Doyle v. Commissioner, 94 Fed. Appx. 949 (3d Cir. 2004); Pierce, 85 T.C.M. (CCH) at 1561, T.C.M. (RIA) at 1009. But see United States v. Evans, 513 F. Supp. 2d 825 (W.D. Tex. 2007), corrected on reconsideration, United States v. Evans, 100 A.F.T.R.2d 2007-6811 (W.D. Tex. 2007). 245. Ohrman, 86 T.C.M. (CCH) at 505, T.C.M. (RIA) at 1668. 246. 96 T.C.M. (CCH) 227, 234, T.C.M. (RIA) ¶ 2008-230 at 1230 (2008). 247. Id. 684 Florid Tax Review [Vol.12:8 v. United States,248 a woman transferred substantially all of her assets to her son when she realized there was tax liability due and her suit to quiet title to her home failed. Considering the requesting spouse’s unclean hands, the court noted in that Wiener the non-requesting spouse transferred the property; in Andrews, the requesting spouse was the transferor. When determining whether a requesting spouse enjoyed a significant benefit from the tax deficiency, courts often ignore the fungibility of money. Instead, courts are likely to find a significant benefit if the government can trace money owed the government to specific expenditures. In Argyle v. Commissioner,249 the IRS traced the nonpayment of tax to the purchase of a new car, and the court agreed that this was a significant benefit. Receiving money on divorce if traceable to the deficiency, even if the marriage was abusive, has been held to be a sufficiently significant benefit.250 On the other hand, in Jones v. U.S.,251 the court held that although there remained some assets that were acquired during the marriage, none were traceable to the understated income and, therefore, the wife did not significantly benefit. In addition, courts rarely look at what would have happened if the taxes had been paid. In Billing v. Commissioner,252 in which a husband sought relief, the court held that it was not a significant benefit to the husband that the couple was able to continue their free-spending lifestyle and afford to purchase a larger house because the wife had spent most of the embezzled money on herself. The court did not consider whether the couple could have afforded the lifestyle or the larger house if the wife had been required to fund her own spending from non-embezzled funds. Applying a mathematical analysis to determine whether the requesting spouse significantly benefited is also unsatisfying as neither Congress nor the Treasury Department has provided a numerical amount that constitutes a significant benefit. Nevertheless, in Haltom v. Commissioner,253 the court applied such an approach. It concluded that of the $275,000 that the husband, as the non-requesting spouse, contributed to family finances from 1990 to 1992, $230,000 benefited the requesting spouse — only $25,000 more than the income they reported. Because this was less than 15 percent of the couple’s adjusted gross income, the requesting spouse was held not to 248. 69 F. Supp. 2d 972, 974 (N.D. Ohio 1999), aff’d, Andrews v. Taylor, 225 F.3d 658 (6th Cir. 2000). 249. T.C. Summ. Op. 2010-129. 250. Estate of Gurr, T.C. Summ. Op. 2002-7. 251. 322 F. Supp. 2d 1024 (D.N.D. 2004). 252. 94 T.C.M. (CCH) 183, 187, T.C.M. (RIA) ¶ 2007-234 at 1427 (2007). See also Beatty v. Commissioner, 93 T.C.M. (CCH) 1422, 1427, T.C.M. (RIA) ¶ 2007-167 at 1123 (2007). But see George v. Commissioner, supra note 234. 253. T.C.M. (RIA) ¶ 2005-209 at 1597. 2012] Empirical Study of Innocent Spouse Relief 685 have significantly benefited and, therefore, the burden was unfairly imposed on her. 2. Crushing Burden a. Economic Hardship In 1998, members of Congress referred to the “financial and emotional distress” imposed on innocent spouses by unrelieved tax burdens.254 Although Congress had been concerned about innocent spouses’ economic hardship, after the enactment of section 6015 the IRS was deluged with many claims for relief by divorced or separated spouses before the IRS had noticed anything wrong with the couples’ returns.255 In 2000, Congress responded by enacting a prohibition on filing claims for relief until a deficiency was asserted.256 Therefore, a requesting spouse must now be assessed, and therefore suffer a tax deficiency, before he or she can seek relief under section 6015(c). This is consistent with the original congressional desire that the tax burden to be relieved should impose true hardship on the requesting spouse. In its interpretation of the statute, the IRS incorporated this congressional objective as one equitable factor for purposes of sections 6015(b) and (f).257 For this purpose, the IRS defines economic hardship as the inability to pay reasonable basic living expenses.258 Determining whether a requesting spouse will suffer economic hardship requires a facts and circumstances test that looks at each requesting spouse’s unique circumstances.259 These circumstances include the requesting spouse’s age; employment status and ability to earn; number of dependents; the amount reasonably necessary for food, clothing, and housing; the cost of living for 254. 144 CONG. REC. S1073 (1998) (statement of Sen. Bob Graham). See also notes 58–68. 255. Robert Steinberg, Three at Bats Against Joint and Several Tax Liability: (1) Innocent Spouse (2) The Election to Limit Liability And (3) Equitable Relief: The Treasury and Courts Begin to Interpret IRC 6015 After Enactment of the IRS Restructuring and Reform Act of 1998, 17 J. AM. ACAD. MATRIM. LAW. 403, 407 (2001) [Hereinafter Steinberg, Three at Bats]; Ryan Donmoyor, ABA Tax Section Meeting: Divorce Lawyers, Tax Lawyers Split on Election of Proportionate Liability, 98 TNT 149-2 (August 4, 1998) [hereinafter Donmoyor, Divorce Lawyers]. 256. Consolidated Appropriations Act 2001, Pub. L. No. 106-554, Appx. G, § 313, 114 Stat. 2763, 2763A640-43 (2000). For debates regarding the value of the protective elections see Donmoyer, Divorce Lawyers, supra note 255; Steinberg, Three at Bats, supra note 255, at 407. 257. Rev. Proc. 2003-61, 2003-2 C.B. 296, 298 258. This definition is made by reference to Reg. § 301.6343-1(b)(4). 259. Reg. § 301.6343-1(b)(4). 686 Florid Tax Review [Vol.12:8 the geographic area; and any extraordinary circumstances. Therefore, when one critic of the current innocent spouse relief stated that failing to find economic hardship is “in essence determining that she would have no trouble paying the tax,” that is not the measure of economic hardship.260 The issue of economic hardship was raised in 187 of the section 6015 cases, but of the seventy-two cases granting relief at least in part on equitable grounds, 22.2 percent, the court did not mention economic hardship.261 However, as shown in the following chart, when judges found economic hardship, it was a strong factor in favor of relief. ECONOMIC HARDSHIP Found Not find IRS conceded Requesting spouse conceded Not enough evidence Neutral or no conclusion Number of cases 40 42 5 9 85 6 Percentage of cases 21.4% 22.5% 2.7% 4.8% 45.5% 3.2% Number of win 37 7 3 2 11 2 Winning percentage 92.5% 16.7% 60.0% 22.2% 12.9% 33.3% In order for a requesting spouse to establish that imposition of liability would create an economic hardship, evidence of the requesting spouse’s financial situation must be presented. In 44.9 percent of the cases raising the issue of economic hardship, the court noted that the requesting spouse did not offer sufficient evidence to support a claim of hardship. It was insufficient when one requesting spouse argued, “It is simply baffling that respondent cannot determine for itself that petitioner would suffer economic hardship if relief from joint and several liability is not granted when it was garnishing $557.45 from her paychecks leaving her a paltry $356.55 for two (2) weeks take home pay.”262 Similarly, hardship was not established when a wife claimed that if all of her assets were liquidated and paid towards the assessment, the couple would still owe more than $1 million in taxes or when another wife argued that the economic hardship factor was “discriminatory and unconstitutional.”263 260. Schumacher, Administrative Process, supra note 98; Commissioner v. Neal, 557 F.3d 1262, 1278 (2009); Alt, 101 Fed. Appx. at 44. 261. Although more cases defined economic hardship under section 6015(f) than under section 6015(b), there was no meaningful difference in the interpretation of the factor if the case was won under one or the other. 262. Ware v. Commissioner, 93 T.C.M. (CCH) 1196, 1198, T.C.M. (RIA) ¶ 2007-112 at 789 (2007). 263. Chou v. Commissioner, 93 T.C.M. (CCH) 1152, 1158, T.C.M. (RIA) ¶ 2007-102 at 732 (2007); Demirjian v. Commissioner, 87 T.C.M. (CCH) 841, 844, 2012] Empirical Study of Innocent Spouse Relief 687 There is, however, no more definitive rule defining economic hardship than that provided by the regulations. The National Taxpayer Advocate reported that 65 percent of those requesting relief make less than $30,000 per year.264 That information is unconfirmed by the opinions. However, some amount of annual earnings might make it impossible to be too heavily burdened. In Feldman v. Commissioner,265 the court found that an attorney earning $130,000 per year was “totally dissimilar from other requesting spouses, who were living at or near poverty level at the time of their request.”266 In Schepers v. Commissioner,267 the requesting spouse complained that he would be required to work until he was seventy-five to pay off the liability. The court dismissed this concern, finding that collection would likely be confined to the ten-year collection period and the requesting spouse would therefore not suffer unduly. The relative inconsistency of the application of the Treasury Department’s standard is illustrated by a comparison of Rice v. Commissioner268 and Stephenson v. Commissioner.269 In Rice, the Court argued that the wife appeared “to have the ability to work more than 20 hours a week and to earn more income,” and she was not granted relief.270 On the other hand, in Stephenson, the wife was held to face economic hardship although she had quit three jobs. Although one can imagine how these cases could be reconciled, the court did not undertake that informative step that would aid the IRS in future applications of the factor. To measure whether an unmitigated tax burden would be crushing to the requesting spouse (or cause economic hardship), courts generally require specific information regarding the requesting spouse’s expenses and income, although these amounts do not always have to be substantiated.271 T.C.M. (RIA) ¶ 2004-22 at 113 (2004). But see Korchack v. Commissioner, 92 T.C.M. (CCH) 199, 217, T.C.M. (RIA) ¶ 2006-185 at 1283 (2006). 264. NTA, 2005 ANNUAL REPORT, supra note 93, at 328. 265. 86 T.C.M. (CCH) 50, 52, T.C.M. (RIA) ¶ 2003-201 at 1092 (2003). 266. Id. 267. 99 T.C.M. (CCH) 1343, 1344, T.C.M. (RIA) ¶ 2010-80 at 500 (2010). 268. T.C. Summ. Op. 2008-69. 269. 101 T.C.M. (CCH) 1048, 1050, T.C.M. (RIA) ¶ 2011-16 at 70 (2011). 270. T.C. Summ. Op. 2008-69. 271. See, e.g., Drayer v. Commissioner, 100 T.C.M. (CCH) 465, 467–68, T.C.M. (RIA) ¶ 2010-257 at 1550 (2010); Kosola v. Commissioner. 99 T.C.M. (CCH) 1141, 1145, T.C.M. (RIA) ¶ 2010-34 at 211 (2010). Although the 1984 version of relief included a new spouse’s income when determining whether a requesting spouse would suffer economic hardship, the provision was omitted in 1998. I.R.C. § 6013(e)(4)(D) (repealed in 1998). In Farmer v. Commissioner, the court held that even though a wife had remarried, there was economic hardship because the wife could not support herself out of her own assets. 93 T.C.M. (CCH) 1052, 1054, T.C.M. (RIA) ¶ 2007-74 at 581 (2007). Similarly, the court would not 688 Florid Tax Review [Vol.12:8 Nevertheless, sufficient proof often requires significant disclosure of personal information. When a requesting wife refused to answer questions about her residence or other assets, it was deemed impossible for her to suffer an economic hardship.272 On the other hand, in one case where abuse was alleged, the court held that when a wife said she did not have the requisite information regarding economic hardship because she did not want to ask her husband, the IRS had an obligation to probe further.273 Despite the facts and circumstances nature of economic hardship, some generalizations can be drawn from the cases regarding what is required to meet this factor for relief. First, the test is personal to the requesting spouse and cannot be claimed by a requesting spouse’s estate.274 On the other hand, the liability need not create economic hardship, a requesting spouse may win relief if the spouse would be in hardship regardless of the liability; however, simply living in a precarious financial situation is insufficient.275 Similarly, a contingent future hardship, even if caused by loss of a job at the IRS, is insufficient to establish an economic hardship as is a difficulty in liquidating one’s assets.276 In Motsko v. Commissioner,277 even though a requesting husband could not use his assets because they were tied up in divorce proceedings, “that does not render them valueless” and they were used to negate economic hardship.278 Declaration of bankruptcy is one possible indicator of economic hardship. allow the IRS to presume that a requesting spouse’s children would continue paying her expenses. Ferrarese v. Commissioner, 84 T.C.M. (CCH) 400, 402, T.C.M. (RIA) ¶ 2002-249 at 1542–43 (2002). 272. D’Aunay v. Commissioner, 87 T.C.M. (CCH) 1134, 1136-37, T.C.M. (RIA) ¶ 2004-79 at 495–96 (2004). 273. Nihiser v. Commissioner, 95 T.C.M. (CCH) 1531, 1532, T.C.M. (RIA) ¶ 2008-135 at 744 (2008). 274. See Jonson v. Commissioner, 118 T.C. 106 (2002), abrogated, Porter v. Commissioner, 132 T.C. 203 (2009). 275. Beatty, 93 T.C. Memo. (CCH) at 1426, T.C. Memo. (RIA) at 1121; Gilliam v. Commissioner, T.C. Summ. Op. 2004-37. 276. Smith v. Commissioner, 98 T.C.M. (CCH) 349, 352, T.C.M. (RIA) ¶ 2009-237 at 1761 (2009); Kalinowski v. Commissioner, 81 T.C.M. (CCH) 1081, 1086, T.C.M. (RIA) ¶ 2001-21 at 173 (2001). 277. 91 T.C.M. (CCH) 711, 714, T.C.M. (RIA) ¶ 2006-17 at 103 (2006). 278. Id. 2012] Empirical Study of Innocent Spouse Relief 689 BANKRUPTCY FILING Husband claimed relief Husband won relief Winning percentage Wife claimed relief Wife won relief Winning percentage Husband filed 6 1 16.7% 19 9 47.4% Wife filed 2 0 0.0% 11 7 63.6% Couple filed 3 1 33.3% 30 13 43.3% Each filed 0 0 0.0% 2 0 0.0% In 42.5 percent of cases, one of the spouses or former spouses had declared bankruptcy, but there was no consistent finding of economic hardship (or significant benefit) in cases mentioning bankruptcy. However, there does appear to be a gender component as wives who claimed relief, whether they or their husbands filed for bankruptcy, did significantly better than when husbands claimed relief and there was a bankruptcy filing. Prior grants of section 6015 relief should help alleviate economic hardship. In forty-three section 6015 cases, the requesting spouse had previously been granted some amount of section 6015 relief. Nevertheless, in 37.2 percent, thus slightly more than the average success rate of innocent spouse cases, the requesting spouse won further relief. Thus, some judges use consideration of economic hardship when determining the equity of granting relief, but when they do so they are defining their own individual sense of when a tax burden will be crushing. For some judges, that there were other avenues for relief for the requesting spouse was sufficient to deny a finding of economic hardship, regardless of whether a requesting spouse wanted to pursue those options. For example, in Rogers v. Commissioner,279 a husband did not receive relief in the Tax Court, which noted that he was also seeking relief in the family court. In Martinez v. Commissioner,280 the court noted the wife’s situation was “highly sympathetic and credible”; however, “if petitioner is truly suffering from economic hardship, or is unable to pay the debt, then she may want to approach the IRS with a request for relief under a different principle, such as an offer-in-compromise or other collection alternative. . . .”281 b. Not Borne by Others In 1998, Representative Nancy Johnson, chairwoman of the House Ways and Means Oversight Subcommittee, said before the hearings, “Where 279. T.C. Summ. Op. 2010-13. See also Schwind v. Commissioner, T.C. Summ. Op. 2008-119; Thompson v. Commissioner, T.C. Summ. Op. 2008-39. 280. T.C. Summ. Op. 2008-165. 281. Id. 690 Florid Tax Review [Vol.12:8 one spouse has fulfilled their full obligation as a wage-earning, tax-paying American, they can be assured they can get complete relief, and we, the rest of the public, will struggle with their non-performing spouse.”282 Thus, at least some in Congress were willing to relieve the requesting spouse of a tax burden with the understanding that the revenue might never be collected. The Treasury Department resisted Johnson’s approach because it would put the federal government in a less favorable position than other creditors who retained joint and several liability.283 It remains to be examined whether judges, recognizing the economic burden potentially placed on requesting spouses, are influenced by the amount of government revenue that might be lost. The Treasury Inspector General for Tax Administration found that for fiscal year 2004, 6,555 requesting spouses were granted relief in the amount of $117.6 million and 10,439 were denied relief of $260.8 million.284 These numbers cannot be confirmed from the sample because not all section 6015 cases report the amount of taxes that may be relieved. Of the 444 cases, it is possible to discern the amount of the tax obligation involved in 352 cases, or 79.1 percent, but it is not always clear if these sums include interest, penalties, or prior payments. AMOUNT OF LIABILITY < $10,000 $10,000 - $50,000 $50,001 - $100,000 $100,001 - $1 million >$1 million Number of cases 113 102 34 74 29 Number won 40 38 9 26 4 Winning percentage 35.4% 37.3% 26.5% 35.1% 13.8% Of the opinions that include reference to how much tax revenue is involved, 32.1 percent of the cases involved amounts less than $10,000 and 61.1 percent involved less than $50,000. Only 8.2 percent of these opinions involved claims with tax bills of over $1 million, and only 13.8 percent of requesting spouses were able to win these high-value cases. Of all of the cases referencing this factor that were won by the requesting spouse, 74.4 percent involved less than $100,000 in taxes. That a large percentage of cases involve relatively smaller sums is surprising because the IRS’s methods for prioritizing collection focus on the aggregate amount of taxes owed, so that smaller revenue amounts get less 282. Oversight Subcommittee, supra note 55, at 32. 283. Id. 284. TIGTA, ACCURATE RELIEF DETERMINATIONS, supra note 97, at 7-8. 2012] Empirical Study of Innocent Spouse Relief 691 attention until interest and penalties accrue.285 Smaller amounts should therefore be a smaller percentage of deficiencies asserted. Nonetheless, innocent spouse relief may be one of the ways these lower-income taxpayers have to contest collection. In Cummings v. Commissioner,286 a divorced woman working two jobs was granted relief from a $506 liability that originated in her husband’s self-employment. She was granted relief based on her economic hardship and the fact that she did not receive a significant benefit from the nonpayment. However, smaller revenue amounts are no guarantee of relief. In Freulich v. Commissioner,287 a widow had income from gambling that generated $606 in taxes. The Tax Court would not grant her relief because the gambling was her income. In yet another case, that $2,300 of taxes owed seemed small in comparison to the requesting wife’s annual income of $46,000 was enough for the court to discount economic hardship.288 As with small revenue claims, it is hard to define what will win a large revenue case. In Chou v. Commissioner,289 the couple was fighting the assessment of tax because, if placed in the earlier year as the IRS claimed, the couple would owe almost $2 million more in alternative minimum tax because of the exercise of stock options that quickly declined in value. The couple lost. Unlike in Chou, in Barranco v. Commissioner290 and Pierce v. Commissioner,291 the couples enjoyed extravagant lifestyles and blatantly abused the tax system. In those cases the couples also lost. For the three cases with more than $1 million of tax liability owed which were won on the merits, all involved requesting wives, none of the wives were found to have knowledge of the deficiency, two arose from investments in tax shelters, two wives had PhDs and one had a masters degree, and in two cases the husband handled the family’s finances but in the other the wife handled them. Except for very large revenue cases, the amount of the liability does not appear to affect the outcome of the case. Similarly, whether the non- requesting spouse would be able or unable to pay the taxes owed was raised in only ninety-seven section 6015 opinions.292 Therefore, in 78.2 percent of all section 6015 cases, no mention was made of the ability to recover from 285. Olson Testifies on Fairness in IRS Enforcement, 2007 TNT 44-28 (Mar. 5, 2007). 286. T.C. Summ. Op. 2007-77. 287. T.C. Summ. Op. 2007-124. 288. Meadows v. Commissioner, T.C. Summ. Op. 2007-42. 289. 93 T.C.M. (CCH) 1152, 1156 T.C.M. (RIA) ¶ 2007-102 at 730 (2007). 290. 85 T.C.M. (CCH) 778, 785, T.C.M. (RIA) ¶ 2003-18 at 70–71 (2003). 291. 85 T.C.M. (CCH) 1553, 1560, T.C.M. (RIA) ¶ 2003-188 at 1008 (2003). 292. In 57% of the cases in which reference was made to the other spouse, the non-requesting spouse was deceased. 692 Florid Tax Review [Vol.12:8 the non-requesting spouse. Of those ninety-seven cases, only seven stated or implied that the non-requesting spouse could pay the tax. In 43.3 percent of the cases in which the court mentioned that the non-requesting spouse was in no financial position to pay the taxes owed, the requesting spouse nevertheless won relief, above the average success rate. Therefore, judges do not seem particularly concerned about the government losing revenue. In some cases, requesting spouses have been relieved of liability not only for taxes attributable to the non-requesting spouse but also for liability on their own earnings or for refunds they have received.293 For example, in Gilbert v. Commissioner,294 a husband was granted section 6015(f) relief from a tax liability attributable to his own earnings because his wife handled the family’s financial affairs and the court found that he had no reason to know that she would not pay the taxes owed. In Yakubik v. Commissioner,295 a husband was granted relief when he did not know of his wife’s embezzled funds. The underreporting of income allowed the couple to claim the earned income tax credit, and relief meant that he was not required to pay back the refund they had received. The amount of refund was larger in Campbell v. Commissioner,296 in which a wife was relieved of liability after her husband settled a $2.8 million liability for $100,000. The couple had received a $314,000 refund that did not have to be repaid. 3. Summary Although Congress was concerned about the IRS unfairly imposing a crushing tax burden on innocent spouses, courts are concerned with only certain features of that burden. Judges are not particularly concerned about the circumstances surrounding the application for innocent spouse relief. Neither what motivated the requesting spouse, the tax issues involved, or (at least for all but the largest tax obligations) the amount of revenue at stake appear to significantly affect how judges rule. There might be some indicators favoring relief, such as the existence of disallowed deductions or unpaid taxes both having a slightly higher than average success rate, and judges are possibly suspicious when a requesting spouse is self-employed, but these factors are not dominant. Judges care more strongly about the amount of knowledge the requesting spouse possessed of the deficiency and whether the requesting spouse benefited from it. However, both knowledge of the liability and 293. But see Freulich, T.C. Summ. Op. 2007-124. 294. T.C. Summ. Op. 2007-16. 295. T.C. Summ. Op. 2008-74. 296. 91 T.C.M. (CCH) 735, 737, T.C.M. (RIA) ¶ 2006-24 at 138 (2006). The house was also in the requesting spouse’s name and the husband’s situation had since improved. 2012] Empirical Study of Innocent Spouse Relief 693 whether a requesting spouse benefited are vaguely defined. Despite the regulations, courts have yet to establish a generally applicable rule to define these factors. What seems extravagant or crushing to one judge might not to another. Thus, judges impose their own interpretation of congressional intent without producing judicial guidelines for what that intent means in practice. D. Characteristics of Non-Requesting Spouse Congress’s 1998 debates regarding innocent spouse relief depicted the “other” spouse as saddling the innocent spouse with unfair and crushing tax burdens.297 This section evaluates whether the non-requesting spouses of those granted relief by the courts should be characterized as such based on the available evidence. 1. Abusive As described in the prior Part, consequences differ significantly if a court finds a requesting spouse signed a return under abuse as opposed to duress.298 If a spouse signs a return under duress, there is no joint return.299 Originally, if a spouse signed as a result of abuse not amounting to duress there was no relief. Despite current attention directed to the problem of domestic violence, little attention was given to it in 1998, but for an amendment late in the legislative process that recognized the problem.300 The IRS subsequently issued guidance providing that if a requesting spouse 297. See notes 58–68. There is nothing to prevent both spouses from seeking innocent spouse relief and allocating liability between them. This was not mentioned in congressional debates. 298. This section does not make any statement about what should constitute abuse or which spouses were actually abused. The former is beyond the scope of this article and the latter determination is too fact specific to be made from the evidence available in the opinions. For more on domestic violence, see Deborah M. Weissman, The Personal is Political — and Economic: Rethinking Domestic Violence, 2007 BYU L. REV. 387 (2007); Michelle Madden Dempsey, What Counts as Domestic Violence: A Conceptual Analysis, 12 WM. & MARY J. WOMEN & L. 301 (2006); Mary Ann Dutton & Lisa Goodman, Coercion in Intimate Partner Violence: Toward a New Conceptualization, 52 SEX ROLES 743 (2005); Emily J. Sack, Battered Women and the State: The Struggle for the Future of Domestic Violence Policy, 2004 WIS. L. REV. 1658 (2004). 299. Rev. Proc. 2003-61, 2003-2 C.B. 296, 297. 300. 144 CONG. REC. S4468, S4500 (1998) (statements of various Senators); Kerns, Duress, supra note 20; Gary M. Fleischman & Jeffrey J. Bryant, A Critique of the Innocent Spouse Equitable Relief Provisions, 90 TNT 1716 (Mar. 19, 2001); Sheryl Stratton & Emily Field, Innocent Spouse Issues Plague Practitioners, IRS, and Courts, 2000 TAX NOTES 115-4 (June 14, 2000). 694 Florid Tax Review [Vol.12:8 can prove to have been abused, the spouse has a stronger case for equitable relief under section 6015(b) and (f) and can overcome actual knowledge of the deficiency under section 6015(c). Nevertheless, for the equitable test as applied in these cases abuse was only one factor among many, and it is not intended be given more weight than other factors.301 Neither the statute nor the regulations define abuse. Some critics claim that because there is no clear standard defining abuse, it is too hard for requesting spouses to prove. While the recent Notice defines abuse broadly, prior to its publication National Taxpayer Advocate Nina Olson denounced the IRS in 2011 for denying an abused woman relief.302 In her opinion, IRS employees who handle these cases demonstrate an unconscionable lack of knowledge about domestic violence.303 The case at issue was Stephenson v. Commissioner,304 in which the IRS granted an ex-wife relief for one year but not for another, focusing on her lack of economic hardship, her receipt of a significant benefit from the unpaid tax, and her knowledge of the deficiency. The court overturned the IRS and extended relief for both years. The court opined that “the verbal abuse turned into physical abuse, and Mr. Stephenson began throwing items at petitioner when he became angry . . . . If petitioner asked what she was signing, Mr. Stephenson made threats of violence or told her she was not intelligent enough to understand. . . .”305 In the court’s and the Taxpayer Advocate’s opinion, the abuse was sufficient to outweigh the other factors and establish equitable relief. The following charts document the number of cases referencing abuse and the number of cases over time. The latter chart looks only at whether the judge found or dismissed the claim of abuse. 301. Rev. Proc. 2003-61, 2003-2 C.B. 296, 298. Arguing that denying relief “effectively reject[s] our allegations of abuse” implicitly suggests that abuse be made a trumping factor, which is not what Congress intended. Schumacher, Administrative Process, supra note 98. The lack of abuse is not meant to weigh against relief, although the IRS has argued unsuccessfully that lack of abuse should weigh against a requesting spouse. See Washington v. Commissioner, 120 T.C. 137 (2003). Abuse and financial control by the nonrequesting spouse has been given greater weight for finding relief for pusposes of section 6015(f). Notice 2012-8, § 4.03(2)(c)(ii), (iv), supra note 9. 302. Fred Stokeld, Taxpayer Advocate Blasts IRS’s Handling of Innocent Spouse Case, 2011 TAX NOTES 16-9 (Jan. 25, 2011); Notice 2012-8, § 4.03(2)(c)(iv), supra note 9. 303. Id. 304. 101 T.C.M. (CCH) 1048, 1049 T.C.M. (RIA) ¶ 2011-16 at 69 (2011). 305. Id. 2012] Empirical Study of Innocent Spouse Relief 695 CLAIMS OF ABUSE Requesting Spouse Claims Abuse Judge notes no abuse Total abuse claims Judge dismisses abuse claims Judge upholds abuse claims Number of cases 93 56 34 22 Taxpayer wins 25 21 5 16 FINDINGS OF ABUSE OVER TIME From the data, judges are cognizant of abuse claims. In 149 cases, courts mentioned abuse and in 62.4 percent of those courts mentioned abuse only to state that it was not alleged in the case. It is also not true that “[t]here are few, if any, decided cases in which abuse was present and the court denied innocent spouse relief.”306 In twelve cases there was at least a mention of abuse but it was unclear from the opinion whether the court concluded that there was abuse for purposes of its section 6015 analysis. In 60.7 percent of the cases in which abuse was alleged, the judge found that there was no abuse, but in 14.7 percent of those cases the requesting spouse was, nonetheless, granted relief. In 27.3 percent of the cases in which the judge found that there was abuse, the judge did not grant relief. One reason judges give for being hesitant to find abuse, particularly mental or emotional abuse, is that a claim of abuse can itself be abused. “We are aware of the danger that requesting spouses, in trying to escape financial liability, may easily exaggerate the level of nonphysical abuse. Innocent- 306. NADLER, INNOCENT SPOUSE RELIEF, supra note 20, at 40. 0 2 4 6 8 10 12 14 16 18 Judge notes no abuse Total claims of abuse 696 Florid Tax Review [Vol.12:8 spouse cases often spring from the dissolution of troubled marriages, and there is an obvious incentive to vilify the nonrequesting spouse.”307 In all the cases where abuse was alleged, all but two involved divorced, separated, or widowed spouses. It is difficult to decipher from the cases what claim or level of abuse is sufficient to outweigh other considerations weighing against relief. Details of abuse are necessary and police involvement preferable, although rarely do the opinions note a significant amount of detail regarding the abuse. In Knorr v. Commissioner,308 the court did not find abuse when the requesting spouse provided only generalized claims of abuse. In Collier v. Commissioner309 the court noted the need for specific details with independent corroboration. On the other hand, in Fox v. Commissioner,310 the court weighed abuse as a positive factor for relief where a police report corroborated the requesting spouse’s claim of assault. Despite courts’ hesitancy to find abuse, in Nihiser v. Commissioner,311 the Tax Court ruled that abuse is not limited to physical abuse and may include verbal and mental abuse. What qualifies as verbal or mental abuse is unclear, although a threat of a voodoo hex will not meet this criteria.312 Similarly, financial irresponsibility, alleged brainwashing, or destroying someone’s credit alone does not amount to abuse.313 In twenty- five cases, requesting spouses claimed to have been the subject of mental or emotional abuse but not physical abuse; the requesting spouse won 32 percent of these cases. However, in only four of them did the court find that there was abuse. There appears to be a gendered component to judges finding abuse, although the sample is small. Once a judge finds abuse, husbands fare better than wives. 307. Nihiser, 95 T.C.M. (CCH) at 1536, T.C.M. (RIA) at 750. 308. T.C.M. (RIA) ¶ 2004-212 at 1324 (2004). 309. 83 T.C.M. (CCH) 1799, 1809, T.C.M. (RIA) ¶ 2002-144 at 908 (2002). 310. 91 T.C.M. (CCH) 731, 734, T.C.M. (RIA) ¶ 2006-22 at 132 (2006). 311. 95 T.C.M. (CCH) 1531, 1536, T.C.M. (RIA) ¶ 2008-135 at 751 (2008). 312. Gilmer v. Commissioner, T.C. Summ. Op. 2007-132. 313. Pugsley v. Commissioner, 100 T.C.M. (CCH) 454, 458, T.C.M. (RIA) ¶ 2010-255 at 1535 (2010); Stolkin, 96 T.C.M. (CCH) at 145, T.C.M. (RIA) at 1102- 03; Pierce, T.C. Summ. Op. 2003-126; Barnes v. Commissioner, T.C.M. (RIA) ¶ 2004-266 at 1624 (2004). 2012] Empirical Study of Innocent Spouse Relief 697 GENDER OF ABUSE CLAIMANTS Claimed abuse Judge found abuse Percentage of claims Spouse won case Winning percentage Wife 51 21 41.2% 15 72.4% Husband 5 1 20.0% 1 100.0% For example, in Schultz v. Commissioner314 and Gilmer v. Commissioner,315 husbands claimed to have been abused. In the former, despite a finding of abuse, the husband was denied relief because the income was attributable to him and “he does not claim that he would have challenged the treatment of any items on the returns.”316 In the latter, the judge did not find abuse a factor favoring relief because the requesting spouse did not corroborate his claim. Not highlighted in the above chart, both spouses claimed abuse in five cases and, in these cases, judges concluded there was no abuse or, if there was abuse, it did not impact their decision. In some cases, judges needed to be convinced that the requesting spouse signed the return only because of the abuse. In Wiksell v. Commissioner,317 the Tax Court agreed that the husband had dragged the wife out of bed and threw her against the wall, that he had held her hair and slammed her head against the wall in front of her children, and that he had consistently intimidated her and the children and belittled and tormented them. Nevertheless, the Ninth Circuit affirmed, repeating the Tax Court’s earlier findings denying relief: “We are simply not convinced that Carpender would not have signed the two returns only because of demands by David. . . . [Specifically, she] failed to establish a nexus between spousal abuse generally and duress in specific instances, the specific instances in this case being Carpender’s signing of these two tax returns.”318 Reliance on an expert alone might be insufficient for this objective, although practitioners expect that it would facilitate relief.319 In Wiksell, as in only one other abuse case, 314. 100 T.C.M. (CCH) 353, 354, T.C.M. (RIA) ¶ 2010-233 at 1385 (2010). 315. T.C. Summ. Op. 2007-132. 316. The husband was granted relief for the portion of the income attributable to his wife. 100 T.C.M. (CCH) at 354, T.C.M. (RIA) at 1386. 317. Wiksell, 67 T.C.M. (CCH) at 2367, T.C.M. (RIA) at 94–485. 318. Wiksell v. Commissioner, 215 F.3d 1335, *3 (9th Cir. 2000), aff’g, Wiksell v. Commissioner, 77 T.C.M. (CCH) 1336, T.C.M. (RIA) ¶ 99,032 (1999). 319. Michael Schlesinger, Obtaining Innocent Spouse Relief in the Face of the Service’s Propensity to Litigate, 109 J. TAX’N 102, 107 (2008). 698 Florid Tax Review [Vol.12:8 the court made mention of the requesting spouse’s expert testimony.320 In these cases the court was not persuaded because the content of the expert’s testimony was found insufficient to prove abuse. Procedurally, abuse in cases when there is a claim for innocent spouse relief raises other concerns that are beyond the scope of this article, but which are likely to decrease the number of appeals to the courts. Congress requires that the IRS contact the non-requesting spouse when it receives an application for innocent spouse relief.321 No exceptions are granted, even for victims of domestic violence.322 During its internal review, the IRS does not disclose personal information that does not relate to the determination of relief. However, if the requesting spouse appeals to the courts, personal information, such as an address, might be disclosed unless a protective order has been issued.323 This issue was not raised in any of the cases in this sample.324 2. Legally Obligated Under State Law In 1998, much of the discussion in Congress focused on divorced spouses and extending relief to these individuals.325 Courts had disallowed taxpayer arguments that for a divorce court to require a spouse to sign a joint return amounted to duress.326 The executive branch similarly did not share 320. 67 T.C.M. (CCH) at 2367-68, T.C.M. (RIA) at 94–486; Collier v. Commissioner, 83 T.C.M. (CCH) 1799, 1809, T.C.M. (RIA) ¶ 2002-144 at 908 (2002). 321. I.R.C. § 6015(h)(2); Reg. § 1.6015-6. 322. The IRS suggests that a spouse who fears abuse write “Potential Domestic Abuse Case” at the top of Form 8857. See Steinberg, Three at Bats, supra note 255, at 412. 323. R. Prac. & P. U.S. Tax Ct. 325; OFFICE OF CHIEF COUNSEL, C.C.N. CC-2005-011, FREQUENTLY ASKED QUESTIONS REGARDING LITIGATION OF SECTION 6015 CASES IN TAX COURT 3-4 (2005). The Tax Court is required to disclose information and provide an opportunity for intervention. I.R.C. § 6015(e)(4). 324. Cases might not have arisen on this issue because of the chilling effect on abused spouses of potential disclosure to abusers. 325. All references in the Senate Finance Committee hearing were to couples whose marriages had come to an end. Finance Committee, supra note 58, at 148. In the House’s hearings, Representative Johnson urged the IRS follow divorce decrees and was “extremely disappointed” that the IRS was not more receptive to the proposal. Oversight Subcommittee, supra note 55, at 20. 326. Steve R. Johnson, The Duress or Deception Defense to Joint and Several Liability, 6 J. TAX PRAC. & PROC. 15 (2004); Price v. Commissioner, 86 T.C.M. (CCH) 203, 204, T.C.M. (RIA) ¶ 2003-226 at 1276 (2003); Berger v. Commissioner, 71 T.C.M. (CCH) 2160, 2172, T.C.M. (RIA) ¶ 96,076 at 96–628 (1996). 2012] Empirical Study of Innocent Spouse Relief 699 Congress’s concern,327 although the IRS later made marital status a factor for determining whether a requesting spouse should be entitled to equitable relief.328 When asked to consider following divorce decrees’ allocations of liability, the Treasury Department worried that the IRS is not a party to divorce proceedings and that there is nothing in a divorce proceeding to protect the government’s interests.329 To allow divorce or separation agreements to allocate liability for federal tax purposes in a way inapplicable to other creditors would allow state law to trump federal revenue collection.330 Some couples proved eager to do so.331 Thus, there was tension between Congress and the executive as to the weight to be given to private agreements. Under pre-1998 law, the Tax Court repeatedly ruled that tax allocation agreements between spouses and former spouses were not binding on the federal courts, but federal courts no longer apply this rule consistently.332 For example, one court dismissed the significance of the couple’s divorce decree, proclaiming in a case otherwise requiring a balancing of factors, “We need not discuss petitioner’s claim regarding the judgment for dissolution of marriage because such a claim is a State matter.”333 On the other hand, finding the apportionment of liability weighed heavily in favor of relief under section 6015, another court ruled, “The most important factor in this case is intervenor’s legal obligation under the North Carolina court’s order to either directly pay the 1998 Federal tax liability or indemnify petitioner for his payment thereof.”334 Although 255 of the couples in which a spouse requested relief, or 59.6 percent, were divorced or divorcing by the time of the trial, only 243 327. See REPORT ON JOINT LIABILITY, supra note 13, at 43. 328. Rev. Proc. 2003-61, 2003-2 C.B. 296, 298. 329. See REPORT ON JOINT LIABILITY, supra note 13, at 43. Because the rights of other creditors remain, following state divorce decrees shifts collections from the federal government to other creditors and often does little to help the requesting spouse. Id. at 41–44. 330. Oversight Subcommittee, supra note 55, at 20 (statements of Donald Lubick and Linda Willis); REPORT ON JOINT LIABILITY, supra note 13, at 28–9. 331. See, e.g., Acoba v. Commissioner, T.C. Summ. Op. 2010-64. 332. Pesch v. Commissionerr, 78 T.C. 100, 129 (1982); Bruner v. Commissioner, 39 T.C. 534, 537 (1962); Neeman v. Commissioner, 13 T.C. 397, 399 (1949), aff’d per curiam 200 F.2d 560 (2d Cir. 1952); Casey v. Commissioner, 12 T.C. 224, 227 (1949); Ballenger v. Commissioner, 14 T.C.M. (CCH) 651, 651, T.C.M. (P-H) ¶ 55,171 at 55–544 (1955); Willis, supra note 55, at 2. 333. Glenn v. Commissioner, T.C. Summ. Op. 2005-127. 334. Gay v. Commissioner, T.C. Summ. Op. 2003-36. See also Bruen v. Commissioner, 98 T.C.M. (CCH) 400, 404–05, T.C.M. (RIA) ¶ 2009-249 at 1834– 36 (2009); Withers v. Commissioner, T.C. Summ. Op. 2010-73. 700 Florid Tax Review [Vol.12:8 had completed their divorce. Of those, only 103 opinions mentioned an allocation between spouses of the tax liability. DIVORCE AND SEPARATION DECREES Number of cases Taxpayer won Winning Percentage Husband made liable 34 17 50.0% Wife made liable 3 1 33.3% Some division of liability 23 10 43.5% Jointly liable 9 3 33.3% Opinion mentions no or ambiguous provision 30 7 23.3% No mention in opinion 140 64 45.7% Other 4 2 50.0% Fifty-eight of the opinions stated that the couples’ divorce decrees allocated at least a portion of the tax liabilities to a non-requesting spouse. In twenty- six of those, or 44.8 percent, the court granted relief consistent with the divorce decree, but none relied solely upon the divorce decree for its determination. One reason for reticence to rely on divorce decrees is that courts share the IRS’s concern that couples will use separation and divorce agreements to take advantage of the tax system.335 In one case a court worried, “In an effort to avoid paying tax liabilities, married taxpayers . . . could structure future payments so that ownership is attributable to the spouse requesting relief under section 6015, while continuing a jointly financed lifestyle.”336 In response to this fear, divorce and separation agreements are given no evidentiary value if the requesting spouse had reason to know the non-requesting spouse would not fulfill the legal obligation.337 In seven of the cases under review, liability was allocated under a divorce decree but the court found that the other spouse was aware at the time of the agreement that the obligation would not be paid. The requesting spouse won none of these cases. In thirteen of the cases both spouses agreed to share liability or the requesting spouse agreed to be solely liable. Nonetheless, in 23.1 percent of those cases, the court granted innocent spouse relief to the liable spouse. For 335. Rev. Proc. 2003-61, 2003-2 C.B. 296, 297; Oversight Committee, supra note 55, at 11; REPORT ON JOINT LIABILITY, supra note 13, at 43. 336. Ordlock v. Commissioner, 126 T.C. 47, 58 (2006). 337. Rev. Proc. 2003-61, 2003-2 C.B. 296, 298. 2012] Empirical Study of Innocent Spouse Relief 701 example, in Maier v. Commissioner,338 the IRS allowed relief despite a divorce agreement providing that the spouses would be jointly liable; the husband intervened but the court had no jurisdiction to hear his claim. In instances where courts granted relief despite agreements to the contrary, the requesting spouse won a better deal than intended by Congress. Although some critics of innocent spouse relief worry that allowing divorced couples this second bite at the allocation apple might be turning the Tax Court into a new divorce court, few cases turned on this issue. Moreover, few cases invoked a family courts’ requirement that spouses sign a return. In the only case in which the issue came up squarely, Bruen v. Commissioner,339 the divorce court had ordered amended returns be filed jointly with each equally responsible. On the returns, the wife wrote “under protest pursuant to Amended Judgement [sic] following Divorce Nisi” above her signature but, according to the court, this was to protest being forced to pay any of the tax liability, not to void the joint return. Filing jointly decreased the couple’s liability by $7,882. Even when cases are not contingent on what the family court requires, courts handling tax matters become embroiled in family affairs. One court complained, “This case arises from a troubled five-year marriage that produced two children, constant bickering, and numerous mutual accusations of wrongdoing. . . . In this case where neither of the main parties is credible, we piece together the fragments of truth as best we can to decide whether she is entitled to relief under section 6015.”340 Even if the court decides to grant innocent spouse relief, a requesting spouse can still be held jointly liable in divorce court if settlements have not yet been finalized.341 In addition to divorce and separation decrees, courts must also examine other state-imposed obligations and means of relief. Community property offers the opportunity for relief or additional liability. The statute provides that section 6015 is to be applied without regard to community 338. 360 F.3d 361, 363–64 (2d Cir. 2004), aff’g, Maier v. Commissioner, 119 T.C. 267 (2002). 339. 98 T.C.M. (CCH) 400, 402, T.C.M. (RIA) ¶ 2009-249 at 1831 (2009). Weight v. Commissioner, 86 T.C.M. (CCH) 98, 99, T.C.M. (RIA) ¶ 2003-214 at 1147 (2003), involved a divorce decree-mandated joint return and one issue was whether the return was timely, but the case was decided on other grounds. In James v. Commissioner, T.C. Summ. Op. 2004-176, the family court required the couple file separate returns. 340. Stergios v. Commissioner, 97 T.C.M. (CCH) 1057, 1057, T.C.M. (RIA) ¶ 2009-15 at 81 (2009). 341. See Melvyn B. Frumkes, Equitable Distribution of Tax Liabilities, 20 J. AM. ACAD. MATRIM. LAW. 179 (2006). See also Williams v. Commissioner, T.C. Summ. Op. 2009-19, in which the court reserved power pending the outcome of the innocent spouse case. 702 Florid Tax Review [Vol.12:8 property law.342 That provides only limited protection as, depending on the state, separate debts can be satisfied from community property. One community property advocate urged that spouses not have separate taxes come out of community property.343 The Treasury Department refused to follow the suggestion,344 and courts have agreed. Although 157 cases arose in community property states, only ten hinged on community property laws, and in all but one the law was interpreted for the benefit of the government. For example, in United States v. Stolle,345 the District Court held that under California community property laws, “community property tax is available to satisfy a debt from either spouse, even if the other spouse is not responsible for the debt.”346 Other state law property rules, such as transferee liability, also apply.347 However, courts have largely dismissed these obligations to the advantage of requesting spouses, so that they provide little revenue for the government.348 In this sample, only two cases involved transferee liability, and both were resolved to the advantage of the requesting spouse. In Jones v. United States,349 a couple had claimed substantial losses as a result of investments in tax shelters that were subsequently disallowed. The District Court ruled that the statute of limitations for transferee liability, one year beyond that of the taxpayer, had lapsed. In United States v. Evans,350 transfers by an executrix of her late husband’s property to her children were set aside as fraudulent conveyances; however, her fiduciary liability was barred by res judicata as a result of an earlier ruling insulating the executrix herself. 3. Intervening Without debate, in 1998 non-requesting spouses were granted the right to participate in the administrative process and to intervene before the Tax Court in innocent spouse cases, although some academics have since 342. I.R.C. § 6015(a). 343. T.D. 9003, supra note 128. 344. Id. Rev. Rul. 2004-74, 2004-2 C.B. 84, 85–86 sets out a five-step process for determining how much of an overpayment the IRS may apply against one spouse’s separate tax liability. 345. 2000 US Dist. LEXIS 5454 at *18 (C.D. Cal. March 15, 2000). 346. Id. 347. Reg. § 1.6015-1(j)(1). 348. Christian, Joint and Several Liability, supra note 20, at 592–77; Beck, Innocent Spouse Problem, supra note 44, at 402–08. 349. 322 F. Supp. 2d at 1026; I.R.C. § 6901(c)(1). 350. 513 F. Supp. 2d at 834. 2012] Empirical Study of Innocent Spouse Relief 703 questioned the wisdom of allowing husbands to intervene.351 More husbands than wives intervene, as shown in the following chart. In the chart, a win by the government against the requesting spouse was counted as a win for the intervening spouse unless the intervening spouse intervened on behalf of the requesting spouse.352 INTERVENORS Total cases Intervened when government conceded Intervened on behalf of requesting spouse Interven or Won Husband as intervenor 61353 20 4 25 Wife as intervenor 21 2 0 13 In eighty-two of the 444 section 6015 cases, or 18.5 percent, the non- requesting spouse intervened; 43.6 percent of the time an intervenor opposed relief, the intervenor won.354 Husbands won 41.0 percent of the time they intervened while wives won 61.9 percent, in part because husbands were much more likely (90.9 percent) to intervene when the government conceded relief to the other spouse. Four husbands (in five cases) intervened on behalf of a former spouse. Excluding cases where the government conceded relief or intervention was on behalf of the requesting spouse, husbands won as intervenors 56.8 percent of the cases. The number of cases with intervenors is limited because section 6015 does not grant the Tax Court jurisdiction over non-requesting spouses’ petitions to review grants of relief by the IRS unless they are appealed by the requesting spouse to the courts. Thus, in all cases where the intervenor sought an appeal, the court denied jurisdiction. However, the IRS cannot 351. I.R.C. § 6015(h)(2), (e)(4); Beck, Failure, supra note 20, at 950. For a discussion of the rights of intervening (and participating) spouses see Rule 325, supra note 295. See also Rev. Proc. 2003-19, 2003-1 C.B. 371. 352. In only three cases did the intervening spouse win without a government victory, and in each of those cases the spouse won the right to intervene. 353. There are three cases for which is it is impossible to confirm if the intervenor intervened on behalf of or against the requesting spouse; however, in one the spouses remained married. 354. These results differ from those found by Trexler, Contesting, supra note 67, who only looks at cases decided on the merits. See text supra page 20. 704 Florid Tax Review [Vol.12:8 settle a case once a requesting spouse files a petition in the Tax Court unless the non-requesting spouse agrees.355 The justification given by the courts for denying jurisdiction (in addition to the lack of statutory authority) is that spouses who file joint returns are jointly and severally liable for the entire liability.356 Therefore, intervenors suffer no actual harm when the other spouse is granted relief. For example, in Maier v. Commissioner,357 when a New York divorce decree stipulated that both spouses remain liable for all taxes due but the IRS granted the wife relief, the court held, “To the extent that petitioner believes that he has suffered an injustice due to a flaw in the controlling statutory provisions, his recourse may be to seek a legislative remedy.”358 In twenty-two of the cases involving intervenors, or 26.8 percent, the IRS had already conceded relief to the requesting spouse by the time of the trial but the intervenor nonetheless appealed. The intervenor lost every time. After Villela-Willcox v. Commissioner,359 it is questionable whether an intervenor can prevail once the government concedes relief, despite the court’s assurance that “[u]nder appropriate circumstances, we would not be reluctant to deny section 6015(c) relief to a requesting spouse if evidence offered by an intervenor, rather than the Commissioner, demonstrated that such relief was unavailable because of the requesting spouse’s ‘actual knowledge’ of ‘the item giving rise to the deficiency.’”360 However, in Villela-Willcox, the court found the intervenor to be the more credible witness and the “intervenor’s evidence shows petitioner’s connection and involvement with intervenor’s participation” in the tax shelter at issue. Nevertheless, the court concluded that although “intervenor’s evidence is persuasive, . . . it is not so compelling to require that the settlement between respondent and petitioner be disregarded.”361 4. Summary Courts generally place less emphasis on the characterization of non- requesting spouses than did Congress, but the weight given to these factors 355. Corson v. Commissioner, 114 T.C. 354, 364–65 (2000). For fiscal year 2010, the non-requesting spouse intervened in ten cases, or 28% of the time. NTA, 2010 ANNUAL REPORT, supra note 17, at 500. 356. See Holloway v. Commissioner, 322 Fed. Appx. 421 (6th Cir. 2008), cert. denied, 129 S. Ct. 1535 (2009); Baranowicz v. Commissioner, 432 F.3d 972 (9th Cir. 2005). 357. 119 T.C. 267, 276, aff’d, Maier, 360 F.3d 361 (2d Cir. 2004). 358. Id. 359. T.C. Summ. Op. 2009-75. 360. Id. 361. Id. 2012] Empirical Study of Innocent Spouse Relief 705 might be increasing. For example, courts are increasingly sensitive to the issue of abuse in claims for innocent spouse relief, although that sensitivity does not mean that courts often find abuse to be a mitigating factor sufficient to outweigh other factors weighing against relief. Similarly, courts often decide cases in a manner consistent with divorce decrees that allocate liability between former spouses but without overtly relying on those allocations. On the other hand, courts are not opposed to intervening spouses, the nefarious husband, unless the IRS concedes relief to the requesting spouse. Therefore, in courts’ evaluation of these factors, consistent with the current Treasury Department guidance, most appear unwilling to give perceived negative features of non-requesting spouses greater weight than other factors that weigh for or against relief. PART IV. CONCLUSION After a decade of fighting, Kathleen Alioto won relief from almost $2 million in taxes. A year earlier, however, another former politician’s wife was not victorious.362 Susan Wilson and her husband, former Arkansas state senator Nick Wilson, were ordered to return a refund they had received. The refund stemmed from $373,089 of illegal kickbacks Nick received when defrauding the federal and state governments and which the couple had reported on their joint tax returns. Once caught, and as part of his sentencing, Nick was required to make restitution. After doing so, the Wilsons claimed a $128,676 refund based on the income they had earlier reported. The government sent the couple a refund check and then demanded its return. Susan sought to keep the refund, claiming she was an innocent spouse. The court did not think the innocent spouse provisions applied in Susan’s case; the provision could not be stretched that far. “That Susan Wilson was not aware of the criminal activity is not relevant.”363 Although the court felt little need to discuss the issue in detail, Susan was still married, had benefited from the refund, and, while her husband might be nefarious, there was nothing to indicate that he had acted nefariously towards her. Therefore, Susan did not fit within the congressional model of an innocent spouse, and she had to repay the refund. When Congress enacted section 6015 in 1998, the primary concern was providing relief to those it deemed “innocent” and, although Congress did not define the term with precision, there was a focus on divorced and separated wives whose husbands had created crushing tax burdens and then unfairly left their wives to pay the bill. It was important to Congress that the 362. United States v. Wilson, No. 4:06cv001628, 2007 U.S. Dist. LEXIS 86900 (E.D. Ark. Nov. 26, 2007); David Firestone, Arkansas Lawmakers Indicted in Vast Corruption Case, N.Y. TIMES, April 28, 1999 at A18. 363. Wilson, 2007 U.S. Dist. LEXIS 86900 at *10. 706 Florid Tax Review [Vol.12:8 tax deficiency resulted from an error created by the non-requesting spouse and that the requesting spouse not have caused the liability.364 Congress did not want innocent spouse relief to become a means by which couples could avoid their “just” taxes.365 The court did not feel Susan fit this mold. The Treasury Department, when interpreting the statute, expanded on this congressional intent, creating a checklist for relief that is often referenced in judicial decisions. That checklist has recently been revised. However, a review of the cases demonstrates that courts apply their own interpretation of congressional intent, only loosely confined by the terms provided by the executive agency. This result occurs despite the fact that the Tax Court, in particular, is being deluged with innocent spouse cases and the court’s individualized evaluation requires fact specific examinations that it might otherwise delegate to the IRS. Thus, in their evaluations of individual claims, courts are using the Treasury Department’s factors to their own ends as they seek to provide relief to divorced or separated wives who were unfairly left oppressive tax burdens by their guilty husbands. Therefore, the gender of the requesting spouse is somewhat important — marital status is more so. That the tax burden be a crushing one is important in the eyes of the courts, but the standard is fuzzily defined. It has been hard for the courts to define clear limits to economic hardship that can be applied with consistency in the innocent spouse context. The same can be said of whether the tax burden is being unfairly imposed on the requesting spouse, but it is important to the courts whether the requesting spouse had knowledge of the tax deficiency. The nature of the non-requesting spouse is less important. Whether the requesting spouse is abused appears to matter to many judges, but the sample is small. Divorce or separation decrees assigning liability can also operate as secondary evidence but will rarely decide a case. Although these guidelines can be gleaned from the cases, in their opinions judges are neither crafting precise definitions of many of these terms nor defining the relative importance of each. Clearer guidance needs to be established by Congress, and definitions by the Treasury Department, to provide more accurate predictive power for individual cases. Despite these difficulties, this article can respond to the fear that the complexities of section 6015 has meant that the innocent spouse regime “degenerate[d] into a global subjective test of whether the spouse seeking relief can move the judge to sympathy.”366 For those claims that are twice denied by the IRS but progress to the courts, the taxpayers who are most likely to win are those for whom Congress intended to provide relief, while spouses like Susan Wilson are often, but not always, denied. 364. See notes 58–68. 365. Id. 366. Beck, Failure, supra note 20, at 942. 2012] Empirical Study of Innocent Spouse Relief 707 The problem remains that litigating relief can be complicated and costly for both taxpayers and the government. As with all equity claims, the factors considered by the courts may be inconsistently applied. As a result, some within Congress are unsatisfied with current relief, or at least the complaints innocent spouse relief receives, and propose further liberalizing section 6015.367 Whether the IRS’s recently liberalized factors for section 6015(f) relief are a response to this new desire or are consistent with Congress’s 1998 intent for that section is left for a later day. Unless clarity of purpose and in operation is provided for in any new legislation, however, a further liberalized law is likely to face the same criticism as the law of the past, regardless of courts’ ability to implement Congress’s desires. 367. Laura Saunders, A New Push to Protect Spouses, WALL ST. J., May 28, 2011, at B9.