The Taxpayer Advocate Service’s Role During an IRS Shutdown

The ongoing lapse in federal government appropriations is now the longest in modern history. On January 15, the IRS released an updated shutdown plan which substantially alters how the ongoing shutdown impacts taxpayers. Regular PT contributor Bob Kamman described the updated plan here; he summarized the original shutdown plan here. The biggest change of plan is the recall of about half the IRS workforce primarily to process tax returns and ensure that refunds will be paid this filing season. But the plan also calls for expanded operations in another important arena: collections. And automated notices of intent to levy are still being issued. Meanwhile, Taxpayer Advocate Service (TAS) employees are largely furloughed and those who are excepted are not permitted to work on individual taxpayer cases. This situation raises important questions about the IRS and Treasury’s interpretation of the Anti-Deficiency Act. What should be TAS’s role in a shutdown, and who will protect taxpayer rights if TAS is not permitted to engage in those activities?  


In a February 2018 PT post, guest bloggeStuart J. Bassin explained the legal framework around a shutdown:

Spending and taxation authority reside in Congress under Article I, Section 8 of the Constitution; the Executive has no independent taxation or spending power. Separately, Article I Section 9 prevents unauthorized spending, providing that “No Money shall be drawn from the Treasury, but in consequence of appropriations made by law.” When a lapse in appropriations occurs, the government necessarily shuts down. Neither the President nor Congress decides to shut down the Government; it happens automatically under the Constitution. 

Statutory law develops the constitutional prohibitions and invokes the criminal law to enforce the prohibition. Under Section 1341(a) to Title 31 of the U.S. Code, federal employees “may not make or authorize an expenditure or obligation” or involve the “government in a contract or obligation” absent a lawful appropriation. Similarly, under Section 1342, federal employees “may not accept voluntary services for the government … except for emergencies involving the safety of human life or the protection of property.” The “emergency” exception is narrowly defined to exclude “ongoing, regular functions of government the suspension of which would not imminently threaten the safety of human life or the protection of property.” Section 1350 makes a knowing violation of either provision a felony punishable by up to two years in prison. 

So, when there is no Congressional appropriation funding the IRS, the Anti-Deficiency Act allows the agency to continue limited operations only “for emergencies involving the safety of human life or the protection of property, albeit with “volunteer” labor.  

This brings us to the Taxpayer Advocate Service. Given the substantial expansion of IRS operations reflected in the revised shutdown plan, it was disappointing to see the role of the Taxpayer Advocate Service remain the same. Bob Kamman notes: 

There are now two “excepted” Category B employees allowed in each local office: The local TA, and either a group manager or a “lead case advocate.”  Their jobs are to “Check mail to comply with the IRS’s requirement to open and process checks during a shutdown while also complying with the statutory requirements that TAS maintain confidential and separate communications with taxpayers and that TAS operate independently of any other IRS office . . .Screen the mail for incoming requests for Taxpayer Assistance Orders and notify the appropriate Business Unit that a request has been made tolling any statute of limitations.” 

It doesn’t sound like they are allowed to answer the phone or work cases.  Protecting IRS integrity doesn’t extend this far? 

Apparently it doesn’t, at least not in the eyes of the IRS. Multiple TAS employees have privately confirmed that they are not permitted to work on taxpayer cases during the shutdown, even where a hardship is present.

TAS plays a crucial role in the tax collection system, as a backstop protecting taxpayers from serious financial hardship. As such, one might think that TAS activities related to taxpayer hardships would be excepted. Certainly taxpayers threatened with homelessness or living without heat or water view their situations as “emergencies involving the safety of human life or the protection of property”. However, the IRS Office of Chief Counsel has interpreted the “protection of property” prong to refer only to government property (Exception Category B of the shutdown plan: “Necessary for the Safety of Human Life or Protection of Government Property”). The NTA strongly disagrees with this interpretation, and she has called on Congress to clarify the statute.  

In her Fiscal Year 2015 Objectives Report to Congress (June 30, 2014), Nina Olson analyzed the Anti-Deficiency Act and raised several concerns with the IRS’s interpretation of its exceptions as applied to TAS. She also compared the IRS’s 2013 shutdown policy to its prior shutdown contingency plan and to the interpretations of other federal agencies and the OMB. It’s illuminating to revisit that report now in light of the original and revised shutdown plans for 2018/2019. I recommend reading the report for the full background and detailed analysis.  

The good news is that the IRS has addressed or partially addressed some of the NTA’s previous concerns. The January 2019 shutdown plan provides for the recall of 1,989 SBSE collection representatives whose duties include releasing liens and levies as required by law. Shockingly, release of levies was not a function performed at all during the 2013 shutdown. The current plan also preserves TAS’s independence by permitting TAS employees to open their own mail, which also was not allowed in 2013. However, the IRS has not changed its interpretation of the “protection of property” exception and it has not conceded that TAS plays an integral role in protecting the integrity of tax collection and in preventing taxpayers from experiencing irreparable harm.  

The limited IRS telephone service provided for in the new shutdown plan has only been open since January 22, and it is not clear yet whether the current operations plan will be sufficient to protect taxpayers against serious harm from unlawful collection actions. Anecdotal reports on the ABA Tax Section’s Low-Income Taxpayer email list suggest that taxpayers and representatives may experience inconsistent treatment when calling in, with some IRS representatives able to accept faxed authorizations and some not, and some representatives claiming they have no ability to address the caller’s hardship. Problems with front-line collection employees are nothing new to taxpayer representatives. One could argue that the sufficiency of the new shutdown plan is not open to debate, as Congress has already determined in enacting sections 7803(c) and 7811 that the NTA and TAS are a necessary safeguard within the system to protect taxpayers from unlawful collection action.  

When the shutdown ends, I anticipate the NTA forcefully renewing her legislative recommendation that Congress 

Clarify that the emergency exception to the Anti-Deficiency Act for the protection of property includes taxpayer property as well as government property. Alternatively, clarify that the National Taxpayer Advocate may incur obligations in advance of appropriations for purposes of assisting taxpayers experiencing an economic hardship within the meaning of IRC § 6343(a)(1)(D) due to an IRS action or inaction, and that the IRS may incur obligations in advance of appropriations for purposes of complying with any Taxpayer Assistance Order issued pursuant to IRC § 7811.

Such clarification would be a relief for taxpayers in a world in which future shutdowns seem inevitable.  

IRS Updates “EZ Answer” Test Procedures for S Cases

At the recent Low-Income Taxpayer Clinic grantee conference, Keith and I were fortunate to hear from a distinguished panel of Tax Court judges discussing practice before the Court. During the panel, Special Trial Judge Leyden was asked about the most surprising thing she has learned since her 2016 transition from practitioner to judge. She commented that she has been dismayed by the sheer volume of cases that are dismissed for the petitioner’s failure to prosecute, and she encouraged participants in the Tax Court’s clinic program to suggest solutions that the Court might consider. The IRS “EZ Answer Test” may be one step towards ameliorating the problem.  


In October of 2017, the IRS began a pilot program called EZ Answer Test. (AP-08-1017-0018, 10/20/17.) This test program allows IRS Counsel in certain offices to answer an S case without waiting to receive the Administrative File, if Counsel believes it does not need the administrative file to answer the petition. (The pilot does not apply to CDP cases.) The filing of an EZ Answer automatically transfers jurisdiction of the case to Appeals. Recently the IRS updated its EZ Answer Test procedures, tightening one of the timeframes and specifying additional procedures. (AP-08-1218-0016, 12/12/18).  

We have discussed issues relating to answers in Tax Court on PT before. In August Keith discussed updated guidance on handling premature petitions, which Bob Kamman had previously highlighted. 

The Court and IRS Chief Counsel would prefer to resolve cases on their merits. Unfortunately, the problem of nonresponsive petitioners stubbornly persists despite the combined efforts of the Court, IRS, and LITCs. As Judge Leyden noted, it is all too common for self-represented petitioners to drop out of their Tax Court cases and become nonresponsive at some point before trial. Sometimes previously-nonresponsive petitioners appear at Calendar Call if the case is not dismissed before that date, but other times petitioners seem to take no action at all after filing their petition.  

One notion that has gained some stakeholder support is that petitioners stand a better chance of remaining engaged if they are contacted early and often after filing their petition. Petitioners who do not hear anything about their case for several months at a time may give up or they may run into problems that make it difficult to remain engaged. For example, in many jurisdictions tenants can be evicted with very little notice. In Pennsylvania the law allows residential leases to provide for zero notice before an eviction action is filed in court. For someone facing a crisis like eviction with very little time to respond, the problem of keeping warm and safe may understandably occupy all of their time and energy. If the tenant has to move, documents may be lost or the taxpayer may not remember to update their address with the Court when they are able to find new housing. In other cases, taxpayers say that they  temporarily dropped off the radar due to health problems which consumed all of their attention. There are many other reasons.  

One of the main reasons for pretrial delays in S cases (between the filing of a petition and the IRS answer, and again from the filing of the answer to when Appeals contacts the taxpayer, and again from the time Appeals sends the case back to Counsel and when Counsel contacts the taxpayer) is that the IRS administrative file on the case must be physically moved from one office to another. Counsel needs the administrative file to Answer the petition. Then, most pro se cases are transferred to the Office of Appeals to attempt settlement. If the case does not settle in Appeals, the file is sent back to Counsel to prepare for trial. I do not know why the process of transferring the administrative file takes as long as it does, but I believe IRS Counsel when they say that they simply cannot get the file quickly. The IRM even has procedures for creating dummy files when the administrative file cannot reach Counsel in time to answer the case. (See IRM (08-09-2011), Dummy File Procedures.)  

I doubt that increased taxpayer engagement was the only motivation behind the IRS’s “EZ Answer Test” program, but I hope the IRS will study taxpayer engagement in conjunction with the program to see if there is any improvement. Whatever the motivation, it is laudable that the IRS is attempting to reduce the time from when a case is filed to when an Appeals employee contacts the taxpayer and substantively engages them in an attempt to settle the case. Faster pretrial timeframes generally help to keep pro se taxpayers engaged, promoting several of the rights in the Taxpayer Bill of Rights including the Right to Quality Service, the Right to Challenge the IRS’s Position and Be Heard, and the Right to Appeal an IRS Decision in an Independent Forum, not to mention (perhaps most important) the Right to Pay No More than the Correct Amount of Tax.  

Annual Low-Income Taxpayer Representation Workshop

For the last several years, the ABA Section of Taxation Pro Bono & Tax Clinics committee has organized a Low-Income Taxpayer Representation Workshop in early December in Washington, D.C. Keith, Les, and I have all been involved with the committee and the workshop at various times over the years. The workshop is a nice opportunity for practitioners interested in low-income taxpayer representation issues to come together for an afternoon of learning and conversation.

This year’s workshop will be on Monday afternoon, December 3, at the D.C. offices of Morgan, Lewis & Bockius LLP. Workshop organizer Caleb Smith has lined up four hours of CLE/CPE with exciting speakers on important topics including section 199A, tax litigation, and collection due process. While the workshop is designed for practioners who represent low-income taxpayers through a Low-Income Tax Clinic or other pro bono program, all are welcome to attend.

Registration is a bargain at $30, reflecting the Tax Section’s commitment to supporting low-income taxpayer representation, and also Morgan Lewis’s generous hosting. (Shout out also to the workshop’s longtime past host McDermott Will & Emery.)

Preliminary Agenda

1:00 p.m. The 2017 Tax Act and Self‐Employed Workers
This panel will discuss new code section 199A as well as tax planning issues for self‐employed and “gig economy” workers in light of the changes made by the 2017 tax act. The panel will also discuss how these changes in the tax law will be reflected during the filing season on the draft 2018 Form 1040.
Moderator: Caleb Smith, Ronald M. Mankoff Tax Clinic, University of Minnesota Law School, Minneapolis, MN
Panelists: Joseph Tiberio, IRS SB/SE, Washington, DC; Caroline Bruckner, American University Kogod School of Business, Washington, DC; Lisa Sperow, Cal Poly Low Income Taxpayer Clinic, San Luis Obispo, CA.

2:00 p.m. Collectability as a Litigation Tool: Settling with DOJ vs. IRS
This panel will discuss taxpayer collectability as a factor in litigation with the Internal Revenue Service in pre‐assessment Tax Court cases, and with the Department of Justice in post-assessment District Court litigation. The panel will discuss differences in the IRS and DOJ approaches on how to treat a taxpayer’s collection potential as a factor in settling cases.
Moderator: Tameka Lester, Georgia State University College of Law, Atlanta, GA
Panelists: Valerie Vlasenko, Agostino & Associates, Hackensack, NJ; Erin Stearns, Director, University of Denver Low Income Taxpayer Clinic, Denver, CO; Carol Koehler Ide, Assistant Chief, Civil Trial Section, Tax Division, U.S. Department of Justice; additional panelists TBA

3:30 p.m. Break (no CLE/CPE)

3:45 p.m. New Trends and Tactics in CDP Litigation
This panel will discuss emerging trends in Collection Due Process litigation, recent precedential court decisions, and identify traps for the unwary.
Moderator: Omeed Firouzi, Christine A. Brunswick Public Service Fellow, Philadelphia Legal Assistance, Philadelphia, PA
Panelists: Keith Fogg, Director Harvard Federal Tax Clinic, Jamaica Plain, MA; Tom Thomas, University of Missouri, Kansas City; Steve Milgrom, Legal Aid Society of San Diego, San Diego, CA; additional panelists TBA

5:15 p.m. Networking Reception (no CLE/CPE)

Tax Court Clarifies Standard of Review in CDP Payment Disputes

Two recent opinions in Melasky v. Commissioner, Docket No. 12777-12L offer a cornucopia of issues for tax procedure watchers to digest. The Melaskys appealed a Collection Due Process (CDP) determination and ultimately lost. On its way to upholding the IRS determination, the Tax Court tackled several unresolved questions in two precedential opinions.  In its preliminary opinion in Melasky (151 T.C. No. 8), the Tax Court decided the standard of review when the dispute concerns the correct application of a payment. This sets the stage for the Court’s opinion on the merits of the Melaskys’ contentions, which we’ll cover in later posts.  

The facts of the case are unusual. The Melaskys hand-delivered a check for $18,000 to their local IRS office on Thursday, January 27, 2011, with direction to apply the check to their 2009 income tax liability. Unfortunately, on Monday, January 31 the IRS issued a notice of levy to the bank on which the check was drawn, which prevented the check from being honored. The IRS applied the bank levy proceeds to an earlier tax year, not 2009. Subsequently a notice of intent to levy was issued for several tax years including 2009. The Melaskys requested a CDP appeal. They argued that the proposed levy should not be sustained as to 2009 because they had no balance for that year after proper application of their payment. 


Before we get to the caselaw, it’s helpful to review a little background. For both lien and levy notices, section 6330(c) sets out the matters to be considered in a CDP hearing: 

(c)Matters considered at hearing. In the case of any hearing conducted under this section—

(1) Requirement of investigation. The appeals officer shall at the hearing obtain verification from the Secretary that the requirements of any applicable law or administrative procedure have been met.

(2) Issues at hearing

(A) In general. The person may raise at the hearing any relevant issue relating to the unpaid tax or the proposed levy, including—

(i) appropriate spousal defenses;

(ii) challenges to the appropriateness of collection actions; and

(iii) offers of collection alternatives, which may include the posting of a bond, the substitution of other assets, an installment agreement, or an offer-in-compromise.

(B) Underlying liability. The person may also raise at the hearing challenges to the existence or amount of the underlying tax liability for any tax period if the person did not receive any statutory notice of deficiency for such tax liability or did not otherwise have an opportunity to dispute such tax liability.

So, issues raised by the taxpayer under 6330(c)(2) fall into two separate buckets: “any relevant issue relating to the unpaid tax” under 6330(c)(2)(A); and “challenges to the existence or amount of the underlying tax liability” under 6330(c)(2)(B).  

Why does it matter which bucket the taxpayer’s argument falls into? It matters for a couple of reasons, one of which is the standard of judicial review. In sections 6330 and 6320 Congress failed to specify the standard that the Tax Court should use in reviewing CDP determinations. However, the legislative history provides clear direction, which the Tax Court recognized and adopted in Goza v. Commissioner, 114 T.C. 176 (2000):

The conferees expect the appeals officer will prepare a written determination addressing the issues presented by the taxpayer and considered at the hearing. … Where the validity of the tax liability was properly at issue in the hearing, and where the determination with regard to the tax liability is part of the appeal, no levy may take place during the pendency of the appeal. The amount of the tax liability will in such cases be reviewed by the appropriate court on a de novo basis. Where the validity of the tax liability is not properly part of the appeal, the taxpayer may challenge the determination of the appeals officer for abuse of discretion. 

(quoting H. Conf. Rept. 105-599, at 266 (1998)). Therefore, whether a taxpayer’s argument falls under 6330(c)(2)(A) or 6330(c)(2)(B) determines the Tax Court’s standard of review. Taxpayers prefer de novo review since there is a relatively lower hurdle for success. Of course, sometimes review of the underlying liability is not available to the taxpayer under the terms of section 6330(c)(2)(B), and in those situations a taxpayer would prefer to have their contentions reviewed for abuse of discretion than not at all.   

This brings us back to the Melaskys. The proper application of the $18,000 taken from the Melaskys’ bank account is disputed, and if the taxpayers’ view prevails they will have no unpaid liability for the 2009 tax year. Does this dispute fall under 6330(c)(2)(A) – “any relevant issue relating to the unpaid tax”? Or does it fall under 6330(c)(2)(B) – a challenge to “the existence or amount of the underlying tax liability”? It cannot be both, since the statute restricts when a taxpayer may challenge the underlying liability.  

A few months ago, Judge Lauber noted in Morgan v. Comm’r, T.C. Memo. 2018-98 that the Tax Court’s caselaw on this issue has been inconsistent, citing cases as far back as 2001. This is a debate that has been brewing for some time.  

It is worth noting that there was no dispute between the parties to the Melasky case over the standard of review. Both parties agreed that for the 2009 liability, the Court’s review should be de novo because “the Melaskys argue that they had no 2009 tax liability.” (Slip op. at 5) This opinion is a good reminder that the Court is not bound by the parties’ views of the law.  

So, the Court considers the question despite nobody asking, and answers it here in a precedential opinion. In his opinion, Judge Holmes cites his earlier case of Kovacevich v. Commissioner, T.C. Memo. 2009-160. It is worth reviewing Kovacevich for a more thorough understanding of Judge Homes’s reasoning.  

Mr. Kovacevich argued that the IRS had not properly applied five payments he’d made by check. In order to parse section 6330 and figure out which category of argument this fit into, the Court first needed to define the term “tax liability.” Judge Holmes looks to the IRC and finds that “[a] tax liability is the tax imposed by the Code on a particular taxpayer for a particular tax year. Sec. 26(b)(1).” With that definition in mind,

challenges to the proper crediting of checks that a taxpayer sends to the IRS are not ‘challenges to the underlying liability,’ because they don’t raise questions of the amount of tax imposed by the Code for a particular tax year. They raise, instead, questions of whether that liability remains unpaid. 

Kovacevich, slip op at 14. In the Kovacevich litigation and in a 2014 Chief Counsel Notice, the IRS agrees, and further argues that this conclusion finds support in the structure of section 6330(c). See Notice CC-2014-002 (May 5, 2014).  

Judge Holmes has not changed his mind since deciding Kovacevich and he reiterates his earlier reasoning and conclusion in Melasky. Despite the agreement of the IRS and the taxpayer, the Melaskys will receive abuse of discretion review for all tax years, including 2009.  

Judge Holmes’s analysis in Kovacevich and Melasky is consistent with the IRS’s views set out in Notice CC-2014-002, except for one point: whether overpayment credits are different from payments such that they fall into a different 6330(c)(2) bucket.  

Melasky did not involve overpayment credits, so the opinion naturally does not analyze the issue. However, Judge Holmes does take pains to distinguish it. The main puzzle for me in both the Kovacevich and Melasky opinions is Judge Holmes’ deferential treatment of Landry v. Commissioner, 116 T.C. 60 (2001). Both opinions distinguish LandryLandry involved a taxpayer who elected to apply his tax refunds to his estimated tax for the following year. Things did not go smoothly for Mr. Landry and he ended up in CDP arguing about his entitlement to some of these credits. The Landry opinion takes just one sentence plus a cite to Goza to conclude that a dispute regarding the amount unpaid after application of overpayment credits places “the validity of the underlying tax liability” at issue. Slip. Op. at 5. There is no analysis of the issue or recognition that some might argue otherwise. It is quite a contrast to the detailed parsing and analysis of the Kovacevich opinion.  

In an income tax case, if the underlying liability is the “tax imposed under subtitle A” of the Code, it is not obvious to me that disputes about overpayment credits are disputes regarding the underlying liability. Under section 6513(d), overpayment credits applied from a prior year return “shall be considered as a payment of the income tax for the succeeding taxable year…” (emphasis added). In contrast, the Code uses different language to describe refundable credits (which are also listed as “payments” on Form 1040, alongside estimated taxes). For example, section 32 begins, “In the case of an eligible individual, there shall be allowed as a credit against the tax imposed by this subtitle for the taxable year…”

The Landry opinion seems to me inconsistent with the reasoning in Kovacevich, and it is plainly inconsistent with the Chief Counsel’s position. Notice CC-2014-002 argues that, “payments and overpayment credits and their proper application have no effect on how much tax is imposed by the Code.” No distinction is made between payments and overpayment credits. This approach seems right to me under the analytical framework used by both the IRS and Judge Holmes. However, I welcome comments on this. There may be Tax Court opinions addressing this issue in depth which I did not find.  

If overpayment credit disputes under section 6513 are subject to de novo review but also the limitations of 6330(c)(2)(B), you could have situations where a taxpayer ends up in very different places depending on whether they elected to have their refund credited to next year’s taxes or whether they got the cash and then paid estimated taxes with it. This does not seem ideal from a taxpayer fairness point of view, since nearly identical taxpayers will get different levels of judicial review. It reminds me of an earlier post, where the distinction between an overpayment credit and a refund made all the difference.  

For now, the Tax Court has decided that payment disputes are not “challenges to the existence or amount of the underlying liability,” but are instead “a relevant issue relating to the unpaid tax.” The law is clearer than it was, although many disputes remain to be litigated.

Third Party Fraud and APTC Repayment Liability

As we move into fall, it’s time for the 10.6 million individuals with “Obamacare” insurance to start thinking about 2019 open enrollment. The Centers for Medicare and Medicaid Services (CMS) released three reports in July illuminating aspects of the Health Insurance exchanges (or Marketplaces) created by the Affordable Care Act (ACA). These reports are described in detail by Katie Keith on the Health Affairs blog. Unsurprisingly, most people with 2018 Marketplace plans are receiving advance premium tax credits (APTC) to subsidize their premiums (87%, up slightly from 84% in 2017). CMS points out that premiums in the exchanges are rising and may be pricing out consumers who do not qualify for APTC.  

The Premium Tax Credit can make health insurance affordable for people without other options, but its structure of advance estimated payments combined with a sheer eligibility cliff when the advance payments are reconciled inevitably leads to harsh outcomes in some cases. (A few were discussed on this blog here.) As APTC absorbs the cost of premium increases, the stakes will only become higher for taxpayers. This blog post gives a brief background on APTC reconciliation in the context of the Tax Court’s deficiency jurisdiction, then highlights one circumstance in which taxpayers should be able to avoid APTC liability: fraudulent enrollment.  


Reconciliation of Advance Premium Tax Credits in the U.S. Tax Court 

The Premium Tax Credit is payable in advance through an ACA exchange, subject to reconciliation on each year’s tax return. See 42 U.S.C. § 18082section 36B(f). (“Marketplace” is the federal government’s preferred term in public communications; however, this blog will follow the statute and regulations in referring to exchanges.) Under section 36B(f), excess APTC awarded by an exchange is considered an income tax liability, subject to certain caps. If a household’s modified adjusted gross income reported on the tax return is above 400% of the federal poverty guideline, the taxpayers must repay all APTC received by themselves or their tax dependents. This eligibility cliff leads to harsh results as many including the National Taxpayer Advocate and the Tax Court have recognized.  

As an income tax liability, a taxpayer’s excess APTC may be redetermined by the Tax Court if the IRS issues a Statutory Notice of Deficiency and the taxpayer timely appeals. See sections 6211 through 6216. However, Tax Court review may not get the taxpayer the result they desire. Some of the most frustrating APTC cases for taxpayers involve government or third-party culpability. For example, in McGuire v. Comm’r, 149 T.C. 9 (2017), the exchange failed to process an income change that the taxpayers duly reported. It erroneously continued APTC payments even though the taxpayers’ income was too high. The Tax Court expressed sympathy but found there was nothing it could do to help the taxpayers avoid repayment, because they had received APTC to which they were not entitled. Likewise, in Gibson v. Comm’r, T.C. Memo. 2017-187, the taxpayers’ young adult dependent had signed up for APTC without the taxpayers’ knowledge. Since they did not disclaim their son as a dependent, the taxpayers were stuck with the repayment obligation.  

The problem for taxpayers hoping to avoid strict reconciliation is that section 36B simply does not have a mechanism to consider equity in the reconciliation of APTC. The U.S. Tax Court was created by Congress, not the U.S. Constitution, and as an “Article I” court its powers are limited to those granted by Congress. See Rawls Trading, L.P. et al. v. Comm’r, 138 T.C. 271, 292section 7442. In a nonprecedential case involving tax treatment of a retirement annuity, Judge Armen provided this explanation with helpful citations: 

Petitioners should understand that the Tax Court is a court of limited jurisdiction and that we are not at liberty to make decisions based solely in equity. See Commissioner v. McCoy, 484 U.S. 3, 7 (1987); Woods v. Commissioner, 92 T.C. 776, 784-787 (1989); Estate of Rosenberg v. Commissioner, 73 T.C. 1014, 1017-1018 (1980); Hays Corp. v. Commissioner, 40 T.C. 436, 442-443 (1963), affd. 331 F.2d 422 (7th Cir. 1964) In other words, absent some constitutional defect, we are constrained to apply the law as written, see Estate of Cowser v. Commissioner, 736 F.2d 1168, 1171-1174 (7th Cir. 1984), affg. 80 T.C. 783 (1983), and we may not rewrite the law because we may deem its “‘effects susceptible of improvement'”, see Commissioner v. Lundy, 516 U.S. 235, 252 (1996) (quoting Badaracco v. Commissioner, 464 U.S. 386, 398 (1984)). Accordingly, petitioners’ appeal must, in this instance, be addressed to their elected representatives. “The proper place for a consideration of petitioner’s complaint is in the halls of Congress, not here.” Hays Corp. v. Commissioner, supra at 443. 

Zedaker v. Comm’r, T.C. Summary Opinion 2011-64. Given the statutory language and the Tax Court’s limited jurisdiction, taxpayers must generally seek remedies elsewhere for inequitable APTC debts.  

Addressing erroneous APTC: the exchange regulations 

A taxpayer who disputes the enrollment or APTC information provided to the IRS by the exchange should try to resolve the dispute with the exchange. It is always a good idea to try to address disputes through the exchange, even if you have a plausible argument in Tax Court. (There have been instances in which the exchange’s Form 1095-A did not match the insurance company records; it might be possible to prevail in Tax Court in such a case.) Exchanges do not like to make retroactive changes, however. After all, the federal government is relying on private insurance companies to offer insurance on the exchanges, and those companies can lose money from retroactive enrollment changes. The exchanges have tried to balance the financial needs of insurance companies with the reality that Form 1095-A can bring genuine errors and other compelling situations to light.  

In narrow circumstances, therefore, third-party misdeeds and exchange errors can entitle a taxpayer to nullify their exchange enrollment and avoid any APTC repayment obligation. Generally, the taxpayer must pursue this through the exchange or through their purported insurance company. Two recent practitioner inquiries reminded me that this is very much a live issue that needs to be identified as soon as possible when a taxpayer seeks assistance. Time is of the essence; the exchange regulations guarantee taxpayers only a short window to request retroactive changes.  

The CMS and Health and Human Services (HHS) exchange regulations at 45 C.F.R. § 155.430(b)(1)(iv) allow enrollees to retroactively cancel coverage when  

(A) The enrollee demonstrates to the Exchange that he or she attempted to terminate his or her coverage or enrollment in a QHP and experienced a technical error that did not allow the enrollee to terminate his or her coverage or enrollment through the Exchange, and requests retroactive termination within 60 days after he or she discovered the technical error. 

(B) The enrollee demonstrates to the Exchange that his or her enrollment in a QHP through the Exchange was unintentional, inadvertent, or erroneous and was the result of the error or misconduct of an officer, employee, or agent of the Exchange or HHS, its instrumentalities, or a non-Exchange entity providing enrollment assistance or conducting enrollment activities. Such enrollee must request cancellation within 60 days of discovering the unintentional, inadvertent, or erroneous enrollment. For purposes of this paragraph (b)(1)(iv)(B), misconduct includes the failure to comply with applicable standards under this part, part 156 of this subchapter, or other applicable Federal or State requirements as determined by the Exchange. 

(C) The enrollee demonstrates to the Exchange that he or she was enrolled in a QHP without his or her knowledge or consent by any third party, including third parties who have no connection with the Exchange, and requests cancellation within 60 days of discovering of the enrollment. 

This right to retroactive cancelation was added to the regulations in the Notice of Benefit and Payment Parameters for 2017, effective May 9, 2016. Note that there is no provision for erroneous APTC: if a taxpayer knowingly enrolled in coverage but received too much APTC, the exchange regulations do not offer a remedy.  

Insurance broker fraud is of particular concern and is a major reason that exchanges grant retroactive cancellations. One of the earliest reported examples came out of North Carolina in 2015. An insurance broker collected names and SSNs at homeless shelters, and ultimately enrolled 600 people. This earned him $9,000 per month in commissions, until the insurance company terminated the relationship. Some of the people enrolled were told they were getting “free insurance”, but others said they did not know they were signing up for insurance at all. The applications conveniently inflated the taxpayers’ income to exactly 100% of the federal poverty level, where they would not owe a monthly premium. While the broker collected his commissions, the enrollees were stuck with insurance that they could not use (for lack of funds to meet the deductible or cost-sharing) or did not know about.  Just having “free” insurance caused hardships for those who relied on programs for the uninsured to receive prescriptions and medical care.  

Reports of broker fraud continued in 2016 and 2017, leading CMS to hold a webinar and issue specific instructions to issuers on July 31, 2017, allowing enrollees meeting certain criteria to fast-track their cancellations. In the webinar slide deck, CMS notes: 

Many of the complainants only learned that they had been enrolled in QHPs when notified by the IRS that their tax refunds would not be processed until they submitted Form 8962 to reconcile their Premium Tax Credit. …Because contact information for consumers may not be correct, the 1095-As did not reach many of the enrollees. 

Also, many of the complaining consumers had other health insurance coverage.  

CMS also issued Examples for Issuers of QHPs in the Exchanges of Elements Demonstrating an Appropriate Rescission, which allows insurance companies to rescind coverage if they suspect fraud and the enrollee either confirms it or cannot be contacted. Finally, CMS’s instructions for broker fraud cases were reiterated on November 20, 2017. Fast-track cancelation is authorized for cases meeting five criteria: 

1. The consumer stated directly to CMS through the FFE Call Center that he/she did not enroll in the Exchange, did not give authorization or consent to an enrollment, and did not want the coverage;

2. The enrollment was completed by an agent or broker or an individual acting under the agent or broker’s direction or control;

3. The consumer is receiving 100% APTC or, if not 100%, the portion of the premium that is the responsibility of the enrollee was not made in whole or in part resulting in the termination of the policy;

4. The issuer has had no contact from the enrollees such as calls to customer service, emails, letters or any other direct contact, with the exception of communications from the enrollee stating that they did not know about or consent to the enrollment; 

5. No claims have been filed for any of the enrollees on each policy.

The regulation permitting cancelation can encompass a broader range of circumstances, but taxpayer representatives should check to see if their client meets the criteria for faster resolution of their dispute.  

One note on timing. While the regulation grants a 60-day dispute window, I would encourage advocates to try for cancelation in compelling situations even if the taxpayer discovered the fraudulent enrollment over 60 days ago. The regulation sets minimum requirements for exchanges to allow cancelations; it does not prevent an exchange from allowing a longer window or from making exceptions to the time limit under a reasoned, consistently applied policy. An exchange’s approach to cases beyond the 60-day window may vary also depending on whether the insurance company consents to the cancelation.  

Final thoughts 

If a taxpayer misses the 60-day dispute window, and the exchange refuses to cancel the coverage, is there any remedy in the Tax Court? As set out above, the taxpayer will likely not prevail by relying on equitable claims or principles. However, a legal argument based on analysis of the Code may have a chance of success. The case of Roberts v. Commissioner, 141 T.C. 569 (2013) (blogged by Scott Schumacher here) may provide some small hope by analogy. In his blog post, Scott explains the legal issue in Roberts: 

Roberts owned several IRA accounts, and during the year at issue, someone withdrew substantial amounts from those accounts.  Roberts said his now ex-wife forged his signature and took the money, while the ex-wife said that she had nothing to do with it.  The IRS took the position that even if she took the money, Roberts as the owner of the retirement accounts, was still taxable on the withdrawals.  

Thanks to the efforts of the University of Washington LITC, the taxpayer prevailed. Scott writes: 

Judge Marvel found that Roberts’ wife had in fact withdrawn the funds from his IRA accounts.  The Court went on to hold that because Roberts did not request, receive, or benefit from the IRA distributions, he was not a payee or distributee within the meaning of section 408(d)(1). 

Could a similar legal argument be made in the case of fraudulent exchange enrollment? If a taxpayer had no idea they were signed up for insurance and they meet all five criteria set out by CMS for expedited cancelation, were there actually “advance payments to [the] taxpayer” made within the meaning of section 36B(f)(2)? This may be ultimately a losing argument, but it could be one worth trying. It seems wisest to continue advocacy with the exchange, issuer and CMS, even if the taxpayer is pursuing an administrative or Tax Court appeal.


Retroactive Math Error Notices May Be on the Horizon

Earlier this week the IRS publicly released a Program Manager Technical Advice (PMTA) memo evaluating the time limits on math error assessment authority. (POSTS-129453-17, 4/10/18) In it, the IRS concludes that it can lawfully make a math error assessment at any time within the assessment statute of limitations. The IRS currently and traditionally uses its math error authority to make corrections during the initial processing of returns. The potential for a change in this practice is the subject of the National Taxpayer Advocate’s 2019 Objectives Report, Area of Focus #7 


Section 6213(b)(1) allows the IRS to make a streamlined assessment without issuing a Notice of Deficiency if the assessment arises from a math or clerical error. This is not as straightforward as it seems. Congress has gradually expanded the definition of a math error in section 6213(g)(2) to encompass situations that do not fit the commonsense definition or a math or clerical error. Math error authority is attractive to Congress and the IRS because theoretically it prevents clearly erroneous spending at a fraction of the cost of a correspondence audit. However, it is not always that simple, as the National Taxpayer Advocate has repeatedly pointed out. Les has discussed the issues on PT here and here 


The April 10th PMTA was written in response to TIGTA Report 2017-40-042 (July 17, 2017). On pages 5-7 of that report, TIGTA dings the IRS for not incorporating PATH Act restrictions into its return processing procedures for the 2016 filing season. Specifically, the PATH Act prohibited taxpayers from claiming certain credits unless the TINs used to qualify for the credits were issued prior to the due date of the return. The IRS has math error authority to reject claims without the required TIN, but it was unable to put math error procedures in place for the PATH Act TIN requirements until the 2017 filing season. 

The IRS points out that the new restrictions were enacted 32 days before the start of the filing season, which was simply not enough time to adjust its processes and inter-agency agreements. The IRS simply did not have TIN issuance dates available during the 2016 filing season. However understandable the situation, the IRS acknowledged that due to the time crunch it issued improper refunds during the 2016 filing season, and it agreed with TIGTA’s recommendation that it take steps to recover the improper refunds. The PMTA is one such step.

Time Limits on Math Error Assessment 

Both the PMTA and the NTA’s response in the 2019 Objectives Report are well worth reading for a deeper understanding of the history of math error assessments and the administrative and policy considerations involved in their use. While the PMTA indicates that the IRS is considering the use of math error authority to recover the specific post-PATH Act refunds identified by TIGTA, its analysis is not limited to that situation.  

The memo points out that section 6213 places no time limits on math error assessments. Likewise, section 6501 setting time limits on assessments does not mention math error assessments or set a different rule for them. Given this statutory structure, and Congress’s repeated expansion of math error authority, the memo concludes that use of math error after return processing is consistent with Congressional intent.

The National Taxpayer Advocate acknowledges that the statute is silent, but she takes issue with the IRS’s ability to significantly expand its use of math error authority after 92 years without explicit Congressional approval or at least a public notice and comment period. She disputes the memo’s interpretation of the 1926 legislative history, since from 1926 to 1976 math error authority was limited to actual arithmetic errors. Congress may never have anticipated that the IRS would change its use of math error procedures so dramatically.  

Finally, the IRS and the National Taxpayer Advocate disagree over whether the use of math error procedures after return processing is necessarily consistent with constitutional due process rights guaranteed by the 5th Amendment. This issue deserves its own post so I will only flag it here. The National Taxpayer Advocate does not claim that retroactive use of math error can never be constitutional, but she urges the IRS to more seriously consider the hurdles and limitations faced by the targeted taxpayers, as well as the importance to the taxpayers of the tax credits being targeted. These factors affect not only policy considerations, she argues, but also constitutional due process analysis.  

Policy and Fairness Concerns 

Erroneous math error assessments abridge taxpayer rights and cause hardships. The NTA does not argue that taxpayers who received erroneous refunds should be allowed to keep them. Her concern is systemic: even a fact as simple as a TIN issuance date is subject to error. The 2019 Objectives Report cites a study from the 2011 Annual Report to Congress in which 55% of TIN-based “math errors” were subsequently reversed at least in part. The July 2017 TIGTA report also raises concerns about the accuracy of the IRS’s TIN issuance data (“The Methodology for Recreating the Individual Taxpayer Identification Number Issuance Date Resulted in Errors).  

The 2019 Objectives Report notes several other concerns: 

As discussed in prior reports, the IRS’s pre-existing [math error authority] raises the following concerns when the resulting assessments are (or may be) erroneous:  

  • The IRS does not try to resolve apparent discrepancies before burdening taxpayers with summary assessments that they are expected to disprove; 

  • IRS communication difficulties, fewer letters (i.e., one math error notice vs. three or more letters from exam), and shorter deadlines (i.e., 60 days vs. more than 120 days in an exam) make it more difficult for taxpayers to respond timely (e.g., because they want to call the IRS to make sure they understand the letter before responding);  

  • Because it is easier to miss math error deadlines, more taxpayers — particularly low income taxpayers — will lose access to the Tax Court; and  

  • Internal Revenue Code § 7605(b) generally prohibits the IRS from examining a return more than once, but the IRS can examine a return after making a math error adjustment. 

(footnotes omitted). The expansion of math error to post-processing situations only heightens these concerns. As the NTA explains,  

Post-processing adjustments make it more difficult for taxpayers to:  

  • Discuss the issue with a preparer who could help them respond;  

  • Access underlying documentation to demonstrate eligibility;  

  • Recall and explain relevant facts;  

  • Return any refunds (or endure an offset) without experiencing an economic hardship; and  

  • Learn how to avoid the problem before the next filing season. 

The PMTA acknowledges that there are legitimate fairness concerns associated with post-processing math error assessments, and notes that the IRS could choose as a matter of policy to take a different approach to the errors identified in the TIGTA report.

Are Math Error Notices Coming Soon for Returns Filed in 2016? 

When the PMTA was written last April, the IRS was contemplating using math error authority to reclaim erroneous credits issued in 2016. For example, a 2014 return filed in spring 2016 with an EIC claim, if some SSNs on the return were issued after the original due date of the 2014 return. This was kosher before the PATH Act but not after. The assessment statute of limitations for that return is coming up next spring, so IRS will need to decide what to do soon, if it has not already decided.  

Before the IRS starts using math error authority 2 ½ years after issuing a refund, I hope it will seriously consider the National Taxpayer Advocate’s concerns and solicit public comments. Many of the taxpayers affected by the PATH Act changes are immigrants. The need for more Spanish-language and other translated correspondence has been a concern since the PATH Act and is just one example of an issue likely to be raised in public comments.  

The erroneous credits flagged by TIGTA primary affect immigrant taxpayers, but the issues raised by the PMTA and the 2019 Objectives Report affect all taxpayers. Once the IRS begins to use math error notices in post-refund situations, further expansion may be irresistible.  

IRS Office of Chief Counsel Gives Direction on Graev Compliance in Litigation

The IRS issued Chief Counsel Notice 2018-006 advising its attorneys how to address compliance with section 6751’s requirement for supervisory approval of penalties in light of the Tax Court’s decision in Graev III. The notice covers a lot of ground in a short number of pages. It reviews several different ways an IRS attorney might encounter Graev issues in litigation and instructs the attorney how to proceed in each situation. I will not review each issue covered by the notice; I encourage readers to read it in full. This blog post discusses two of the items mentioned in Notice 2018-006: the application of section 6751 to the trust fund recovery penalty and the exception in subsection 6751(b)(2) for penalties automatically calculated through electronic means. 


IRS asserts supervisory approval is not required before imposition of a Trust Fund Recovery Penalty 

Section 6751(b) begins, “No penalty under this title shall be assessed unless…” In a brief paragraph, Chief Counsel Notice 2018-006 directs IRS attorneys to argue that the trust fund recovery penalty (TFRP) under section 6672 is a tax, rather than a penalty. Therefore, the argument goes, supervisory approval for assertion of a TFRP is not required by section 6751. The Notice also flags this issue generally, noting “there may be other taxes that taxpayers will contend are penalties.”  

In support of the IRS position the Notice cites United States v. Rozbruch, 28 F. Supp. 3d 256 (S.D.N.Y. 2014), aff’d on other grounds, 621 Fed. Appx. 77 (2d Cir. 2015). In Rozbruch the government had filed suit under section 7403 to reduce its tax assessments to judgment and to foreclose on property owned by the Rozbruchs. The taxpayers defended in part by arguing that the TFRP assessments were invalid for failure to comply with section 6751’s supervisory approval requirement.  

The Rozbruchs pointed out that section 6672 explicitly calls the TFRP a penalty. Penalties may not be imposed on top of the TFRP. (I have not read the case documents apart from the court’s decision; the taxpayers may have made additional points not mentioned in the decision.) On the other side, the government pointed out that individuals who make payments towards a TFRP are really paying the underlying trust fund taxes; liability for a TFRP is co-extensive with the employment taxes that remain to be paid over. The district court in Rozbruch agreed with the government, citing an Eighth Circuit decision and several Second Circuit district court cases finding the TFRP to be a tax.  

However, the government’s position is not yet established in the Tax Court. The court declined to decide the issue in its April 2018 decision in the CDP case of Blackburn v. Commissioner, 150 T.C. No. 9. (The case was first noted on PT in a Designated Order post by Professor Caleb Smith.) In Blackburn Judge Goeke found that it was unnecessary to decide whether supervisory approval is required for TFRPs because the IRS had satisfied the requirement in any event. The IRS had a Form 4183, Recommendation re: Trust Fund Recovery Penalty Assessment, bearing the approval of the recommending Revenue Officer’s immediate supervisor, the Acting Group Manager. This was enough to support the Settlement Officer’s finding under section 6330(c)(1) that IRS had complied with administrative procedures in making the assessment. Judge Goeke rejected the taxpayer’s argument that the Settlement Officer should have looked beyond the form to investigate whether the supervisor’s review was “meaningful.” As the opinion points out, Tax Court “caselaw acknowledges that reliance upon standard administrative records is acceptable to verify assessments.” Blackburn, slip op. at 10 (citing Nestor v. Commissioner, 118 T.C. 162, 266 (2002); Davis v. Commissioner, 115 T.C. 35, 41 (2000)).  

Judge Holmes also raised the issue in a TFRP CDP case, Humiston v. CommissionerDocket No. 25787-16 L. In Humiston, no Form 4183 was mentioned in the Settlement Officer’s determination or apparent from the record before the court. In a May 24 order denying the IRS’s motion for summary judgment, Judge Holmes notes that the TFRPs “are called penalties under the Code,” and determines that the novelty of the legal issue weighs in favor of permitting petitioner to raise it at calendar call. I do not know whether petitioner did so, but he now has counsel so perhaps the case will result in a decision on the issue. It appears that case would be appealable to the 2d Circuit Court of Appeals, which of course started all this with Chai 

Next up, Judge Lauber may have the intersection of 6751 and the TFRP before him in the Florida case of Kane v. CommissionerDocket No. 10988-17 L (hat tip: Lew Taishoff). Under a recent order, the IRS has until June 28 to file “a Form 4183 or any other relevant documentation concerning supervisory approval of the TFRPs in question.”  

The way a provision is titled or described does not trump its function for constitutional purposes, but it does provide evidence of Congressional intent for purposes of statutory construction. See National Federation of Independent Business v. Sibelius, 567 U.S. 519, 543-546, slip op. at 12-13 (2012) (finding the individual shared responsibility provision a penalty for purposes of the Anti-Injunction Act but a tax for purposes of constitutional validity). Taxpayers arguing this issue before the Tax Court will need to address both form and function of the TFRP.  

Uncertainty continues around penalties automatically calculated through electronic means 

Readers of this blog know that penalties which are “automatically calculated through electronic means” do not require supervisory approval under section 6751(b)(2). For background, see my previous post on the issue. Chief Counsel Notice 2018-006 largely repeats the position taken in prior IRS guidance (linked in the prior post). The 2018 Notice says, in part:  

Penalties appearing in a statutory notice of deficiency as a result of programs such as the Automated Underreporter (AUR) and the Combined Annual Wage Reporting Automated programs will fall within the exception for penalties automatically calculated through electronic means if no one submits any response to the notice, such as a CP2000, proposing a penalty. However, if the taxpayer submits a response, written or otherwise, that challenges a proposed penalty, or the amount of tax to which a proposed penalty is attributable, then the immediate supervisor of the Service employee considering the response should provide written supervisory approval prior to the issuance of any statutory notice of deficiency that includes the penalty. A penalty is no longer automated once a Service employee makes an independent determination to pursue a penalty or to pursue adjustments to tax to which a penalty is attributable.  

This is in line with prior IRS guidance but it does not do a whole lot to clear up unresolved issues. For one thing, the notice does not mention correspondence examinations. It is unclear whether we should read anything into that. Perhaps the IRS is not ready to publicly take a consistent position on correspondence examinations, but it is a shame the issue was not explicitly addressed in Notice 2018-006. The Notice also fails to clarify the issue of timing, and when precisely a taxpayer’s response arrives too late take a penalty out of section 6751(b)(2)’s ambit.   

The prior post on section 6751(b)(2) examined a nondesignated order in Triggs v. Commissioner. Triggs involves a correspondence examination where the taxpayer did not respond during the audit. The IRS argued that the correspondence exam function had automatically asserted the penalties according to its computer programming without the independent intervention of any human IRS employee, and therefore no supervisory approval was required for either the negligence penalty or the substantial understatement penalty. Although penalty assertion in correspondence examinations may be fully automated in practice, several IRM provisions appear to conflict with this view and require examiners to exercise responsibility for penalty assertions. On April 5, 2018 Judge Leyden ordered further submissions to address these IRM provisions.  

The Triggs case is still pending. The IRS filed its supplement in response to the April 5 order, and subsequently Judge Leyden gave the petitioner until June 29 to respond. The court also provided him with a list of local low-income taxpayer clinics. Unfortunately, to date Mr. Triggs remains pro se. 

IRM provisions also muddy the waters for Judge Lauber in the case of Bowse v. CommissionerBowse has a fact pattern similar to Triggs but in it arises from the AUR program rather than correspondence exam. It also deals solely with the substantial understatement penalty rather than both negligence and understatement. Like Triggs, the Bowse case was brought to my attention by Carl Smith.  

In Bowse, Judge Lauber highlights the timing problems that muddle the application of section 6751(b)(2) to a deficiency case in the Tax Court. Remember, it is the “initial determination” of a penalty “assessment” that requires supervisory approval under section 6751. Professor Bryan Camp has an excellent post explaining why the statutory language does not make much sense and inevitably causes courts to engage in interpretive gymnastics. We can clearly see this in Bowse 

Mr. Bowse and his wife Ms. Vaes were picked up by the AUR program, but they did not respond to the IRS’s letters until after the AUR program issued them a statutory notice of deficiency (SNOD). The SNOD included a 20% substantial understatement penalty. At that point, the taxpayers submitted an amended return, and in response the Office of Appeals reduced the proposed deficiency and understatement penalty amounts. The penalty was still 20% of the proposed deficiency; it was reduced in proportion to the deficiency amount. The taxpayer then appealed the SNOD to the Tax Court.  

On those facts, what constitutes the “initial determination” of the penalty “assessment”? Was it the automated proposal by AUR in the SNOD? Or did the individualized review by Appeals take the case out of 6751(b)(2) territory? Does it matter that the proposed penalty amount changed? To answer this question, one needs to consider both what “initial determination” means and what qualifies as the “assessment” under section 6751.  

As Professor Camp explains, in Graev III the Tax Court interpreted the word “assessment” to essentially mean “penalty assertion.” And so Graev III examines whether certain IRS employees had “authority to make the initial determination of a penalty,” among other issues. 149 T.C. No. 23, slip op. at 17. But before Graev III the word “assessment” in section 6751 was taken at its ordinary meaning in the tax context – the official recording of a liability. Because of this, pre-Graev IRS guidance and IRMs are not wholly consistent with the IRS’s current litigating position. The IRMs in particular raise questions for Judge Lauber, as they did for Judge Leyden in Triggs. 

In Bowse, the IRS argued that no supervisory approval was required because the AUR program had automatically proposed the penalty in its SNOD without any human employee review. However, Judge Lauber is not sure it’s that simple and requests further briefing from the parties on several points.  

In his order Judge Lauber quotes several IRM sections, including this one: 

‘…However, if a taxpayer responds either to the initial letter proposing a penalty or to the notice of deficiency that the program automatically issues, an IRS employee will have to consider the taxpayer’s response. Therefore, the IRS employee will have to make an independent determination as to whether the response provides a basis upon which the taxpayer may avoid the penalty. The employee’s independent determination of whether the penalty is appropriate means the penalty is not automatically calculated through electronic means. Accordingly, IRC 6751(b)(1) requires managerial written approval of an employee’s determination to assert the penalty.’ IRM pt. (Jan. 24, 2012); see also IRM pt. and (3) (Sept. 1, 2012).  

(Emphasis added.) The IRM clearly suggests that the taxpayer can escape automatic penalty calculation by responding to the AUR program’s SNOD. Judge Lauber asks for briefing to address the following five questions: 

(1) By filing a Form 1040X for 2014 after receiving the notice of deficiency, did petitioners “respond * * * to the notice of deficiency that the [AUR] program automatically issue[d],” within the meaning of the IRM provision quoted above?

(2) If so, after considering petitioners’ response, did an IRS employee “make an independent determination as to whether the response provides a basis upon which the taxpayer may avoid the penalty,” within the meaning of the IRM provision quoted above?

(3) By accepting petitioners’ amended return and reducing the deficiency and penalty amounts, did the IRS made a new “initial determination” of the assessment of the penalty?

(4) If so, was that new initial determination of the penalty “automatically calculated through electronic means”?

(5) If not, what IRS officer made the “initial determination” of the reduced penalty, and who was the immediate supervisor of that individual? 

This order is interesting coming from Judge Lauber, whose concurrence in Graev III suggests a straightforward approach. He describes the majority opinion as “requiring supervisory approval the first time an IRS official introduces the penalty into the conversation.” Graev III, Lauber, J., concurring, slip op. at 27. That reading of “initial determination” appears narrower than the IRM’s, since under the IRM a taxpayer’s response to a SNOD can take a penalty out of section 6751(b)(2)’s ambit.  

I suspect the conflict between the IRS’s litigating position and the IRM is temporary. The IRMs cited by Judges Lauber and Leyden (like the 2002 Service Center Advice) reflect the IRS’s pre-Graev III understanding of what “assessment” means in section 6751. If assessment meant the official recording of the liability, then naturally a taxpayer’s response after an AUR SNOD but prior to assessment could take the final penalty decision out of the computer’s hands. Now that penalty “assessment” means penalty “proposal,” though, I would expect to see the IRMs on section 6751(b)(2) changing to remove references to taxpayers responding to the SNOD. It is somewhat puzzling, however, that Chief Counsel Notice 2018-006 does not explicitly mention this change or clarify whether, in the IRS’s view, the SNOD is the cut-off point for taxpayers to respond to an automated penalty proposal in order to trigger supervisory review.  


Insolvency Exclusion for Canceled Debt Nixed by Nominee Bank Account

I have been thinking about alter ego, nominee, and transferee issues recently. At the ABA Tax Section 2018 May Meeting I moderated a panel discussing these issues in the tax lien context. (Materials here, membership or meeting registration may be required.) Normally these topics come up with taxpayers while we are preparing for an offer in compromise, installment agreement, or innocent spouse request. The recent case of Hamilton v. Commissioner, T.C. Memo. 2018-62, reminds us that the nominee doctrine can also be relevant outside of the collection setting.



Mr. Hamilton took out student loans to finance his son Andrew’s education, as many parents do. Sadly Mr. Hamilton later injured his back and became permanently disabled. (We have little information about Andrew, so I like to think that he graduated and used his education to succeed in his chosen career.) Due to Mr. Hamilton’s disability, his lenders discharged his student loans in 2011.

Unfortunately Mr. Hamilton’s problems went beyond his physical disability, and he began spending money erratically. As a result, his wife took over managing their finances. Mr. Hamilton had a large amount of cash, which was placed in Andrew’s savings account on April 1, 2011. We do not know the thinking behind this move, but perhaps Mr. Hamilton felt better about placing his money with Andrew for safekeeping than about placing it with his wife. After April 1, Andrew allowed Mrs. Hamilton to use his online banking credentials and to freely transfer funds from his savings account back to the petitioners’ joint account. Mrs. Hamilton regularly did this, and she paid household bills from the joint account.

Next we come to the Hamiltons’ 2011 joint tax return. Canceled debt is normally gross income under Code section 61(a)(12). However, section 108 lists several exclusions including the insolvency exclusion in section 108(a)(1)(B). It is not always easy for taxpayers to win an exclusion. A recent Designated Order post by William Schmidt includes two cases in which the taxpayers failed to prove that canceled debt should not be income to them. William also discusses the mechanics of claiming the insolvency exclusion.

Insolvency is defined in section 108(d)(3) as “the excess of liabilities over the fair market value of assets,” determined just before the discharge. The amount of canceled debt excluded from income can’t exceed the amount of insolvency. IRC § 108(a)(3). So, a taxpayer with total assets worth $3,000 and total liabilities of $6,000 could exclude up to $3,000 of canceled debt from her income under section 108(a)(1)(B).

One question for the Hamiltons, then, was the fair market value of Mr. Hamilton’s assets immediately before the discharge. The Hamiltons must have transferred the cash to Andrew’s account before the student loans were discharged, because they excluded that cash from their insolvency calculations. On their 2011 tax return the Hamiltons claimed they were insolvent by more than the amount of discharged debt, and therefore they excluded all the canceled debt from their income. Mr. Hamilton would have been solvent if the cash placed in Andrew’s account had been included in the equation.

The IRS examined the return, disagreed with the insolvency conclusion in a notice of deficiency, and the Hamiltons appealed to the Tax Court.

On a side note – readers may be wondering if the 2017 Tax Act would change the analysis in this case. Student loans discharged in 2018 through 2025 due to the death or total and permanent disability of the student are not includable in gross income. See code section 108(f)(5). Unfortunately, Mr. Hamilton was not the student (Andrew was) so I do not think this case would come out any differently if it involved tax year 2018.

In the Tax Court, the IRS argued that the cash in question was still Mr. Hamilton’s property after he placed it into Andrew’s bank account and that Andrew merely owned the account as nominee for Mr. Hamilton. Therefore, the IRS argued that Mr. Hamilton was solvent immediately before the student loans were canceled and therefore he owed income tax on the canceled debt.

The nominee doctrine is really common sense. It applies where a transfer occurs in name only, and the transferor retains beneficial ownership of the property. (Keith previously discussed nominee liens here; guest blogger A. Lavar Taylor discussed purported nominees’ CDP rights in a two-part series here and here.) The Internal Revenue Manual has a summary of the doctrine and it details how the IRS will go about enforcement action when a delinquent taxpayer’s assets are held by a nominee. IRM section explains the theory:

The nominee theory is based on the premise that the taxpayer ultimately retains the benefit, use, or control over property that was allegedly transferred to a third party. Thus, the nominee theory focuses on the relationship between the taxpayer and the transferred property. A transfer of legal title may or may not have occurred, but the government does not believe a substantive transfer of control over the property in fact occurred.

Courts generally look to state law to determine a taxpayer’s property rights, including whether property belongs to the taxpayer under the nominee doctrine. See Fourth Inv. LP v. United States, 720 F.3d 1058, 1066 (9th Cir. 2013). These factors vary somewhat in their wording from state to state but they are remarkably consistent.

In Mr. Hamilton’s case, the Tax Court looked at the nominee factors under Utah law:

(i) the taxpayer exercises dominion and control over the property while the property is in the nominee’s name; (ii) the nominee paid little or no consideration for the property; (iii) the taxpayer placed the property in the nominee’s name in anticipation of a liability or lawsuit; (iv) a close relationship exists between the taxpayer and the nominee; (v) the taxpayer continues to enjoy the benefits of the property while it is in the nominee’s name; and (vi) the conveyance to the nominee is not recorded.

In the recently-blogged Kraus case, the court looked to the nominee factors found in Washington State law:

(1) Whether the nominee paid no or inadequate consideration…; (2) Whether the property was placed in the name of the nominee in anticipation of litigation or liabilities; (3) Whether there is a close relationship between the transferor and the nominee; (4) Whether the parties to the transfer failed to record the conveyance; (5) Whether the transferor retained possession; and (6) Whether the transferor continues to enjoy the benefits of the transferred property.

The Utah and Washington nominee factors are quite similar to the nonexclusive factors listed in the IRM at (3):

a. The taxpayer previously owned the property.

b. The nominee paid little or no consideration for the property.

c. The taxpayer retains possession or control of the property.

d. The taxpayer continues to use and enjoy the property conveyed just as the taxpayer had before such conveyance.

e. The taxpayer pays all or most of the expenses of the property.

f. The conveyance was for tax avoidance purposes.

From these examples one can appreciate the flexibility of the common law as well as its ultimate convergence with common sense (at least in this instance).

In Mr. Hamilton’s case, the Tax Court easily concluded that the transferred funds still belonged to Mr. Hamilton, as Andrew was only holding on to them as Mr. Hamilton’s nominee. The Hamiltons retained full use of the funds. I think this is cleary the right result. The April 1 transfer may have been for the legitimate purpose of protecting Mr. Hamilton’s cash from his erratic spending impulses, but the Hamiltons should not have pushed their luck by excluding the account as an asset in their insolvency calculations.

This case illustrates the importance of asking about potential nominee property in all contexts where a taxpayer’s assets are relevant to a tax issue. Also, it provides an example of a complicating situation that could be easily missed if a tax preparer does not take (or have) time to conduct a thorough interview.

Low-income taxpayer advocates have long complained that free tax assistance sites are not permitted to help taxpayers claim the insolvency exclusion. Taxpayers who have no money to pay a preparer are often forced to borrow to come up with the funds, or else try to self-prepare, which can result in a controversy case when they cannot figure out how to properly exclude the income. This issue was highlighted by the National Taxpayer Advocate in her 2017 Annual Report to Congress, in Most Serious Problem No. 11. The IRS could delineate appropriate boundaries and provide worksheets to facilitate the insolvency determination by VITA, TCE, and AARP assisters and it would be a significant help to many taxpayers. I do not think the Hamilton case detracts from that argument too much. Most taxpayers whose student loans are canceled due to disability do not have $300,000 in cash to worry about. Nevertheless, the Hamilton case illustrates how the insolvency determination is not always straightforward.