Search Results for: graev

Changing a Penalty – Graev Effect

In Castro v. Commissioner, TC Memo 2022-120 petitioner sought to strike an IRC 6662 accuracy related penalty for failure to meet the requirements of IRC 6751(b).  The Court determined that the manager’s approval met the statutory requirements.  The decision here affirms prior case law holding that the Court will not look behind the signature just as it does not look behind a notice of deficiency.  In effect, the Graev test has a Greenberg Express overlay.  Signing the right form at the right time is the key to success for the IRS not proving that it had a good reason for signing the form.


The manager signed the 30-day letter.  It approved the imposition of an accuracy related penalty.  Later, the IRS revised the penalties.  In doing so a manager approved the change, but the Castros complained that by the time of the change the original penalty proposal had been communicated.  The parties submitted his case fully stipulated under Rule 122 after resolution of all issues except the penalty.  As we have discussed before, this rule allows the parties to avoid the time consuming and messy trial if they can agree to all of the facts necessary for the Court to reach its opinion.

As a result of review of the original revenue agent’s report, the report was revised:

On October 18, 2016, GM Relf signed a Civil Penalty Approval Form with respect to the years at issue. In that form, under the “Assert Penalty” heading, an “X” was marked under the “Yes” column for an addition to tax under section 6651(a)(1) and an accuracy-related penalty for substantial understatement of income tax under section 6662(d) for each year at issue.4 Conversely, under the “Assert Penalty” heading, an [*4] “X” was marked under the “No” column for additions to tax under sections 6651(a)(2) and 6654 for the years at issue.

On October 26, 2016, respondent issued the notice of deficiency underlying this case. The penalties and additions to tax determined in the notice of deficiency corresponded to the penalties and amounts shown in the corrected RAR. Specifically, respondent determined only additions to tax under section 6651(a)(1) and accuracy-related penalties under section 6662(a) (not additions to tax under sections 6651(a)(2) and 6654) for each of the years at issue. The amount of each accuracy-related penalty under section 6662(a), and of each addition to tax under section 6651(a)(1) that had been asserted in the original RAR dated July 6, 2016, was also revised downward to reflect the amount shown in the corrected RAR dated October 5, 2016.

The Tax Court has prior precedent holding that the confusing language in IRC 6751(b) regarding “the ‘initial determination’ of a penalty assessment … is embodied in the document by which the Examination Division formally notifies the taxpayer, in writing, that it has completed its work and made an unequivocal decision to assert penalties.”  That document is usually the 30-day letter.  Here, the original 30-day letter does not match the revised one or the notice of deficiency.  Here, the manager approved the 30-day letter by signing the letter transmitting the report to them.  This was insufficient according to the Castros because other evidence suggests that the manager did not really review the penalty until later:

on the totality of the stipulated record, the signature found on the Letter 950 does not demonstrate written managerial approval of the contents of the enclosed RAR, including the asserted accuracy-related penalties. Specifically, petitioners note that the Letter 950 was mailed to them on July 8, 2016, with an enclosed RAR asserting penalties or additions to tax under four distinct Code provisions. The stipulated record, petitioners argue, proves that the first review of the penalties and additions to tax asserted in the RAR did not occur until August 30, 2016, and, once that occurred, GM Relf found errors with respect to the penalties and additions to tax. Accordingly, petitioners argue that when GM Relf signed the Civil Penalty Approval Form on October 18, 2016, he expressly disapproved of two additions to tax that were purportedly the subject of his approval when he signed the Letter 950 on July 8, 2016, and he approved revised amounts of the other penalties and additions to tax. This later and express disapproval, petitioners contend, is the type of contrary evidence that precludes a conclusion that GM Relf’s signature on the Letter 950 approved the contents of the RAR. On the contrary, petitioners insist that the “most logical conclusion [to be drawn from the stipulated record] is that the group manager did not view the Letter 950 as an approving document and did not view the signing of the Letter 950 as an event which triggered his responsibility to review the penalties.”

The Court finds for the IRS because the evidence showed that the manager signed the initial correspondence.  Basically, the Court harks back to earlier cases in which taxpayers sought to have the court look behind the approval.  The Court has consistently rejected efforts by petitioners to have it look behind the approval to see if the manager made a good decision or an informed decision.  It simply looks to see if the manager signed.  If the manager signs the approval form with his or her eyes closed, that will be good enough. 

If the Court must gauge the depth of a manager’s understanding of the penalty imposition, it would be even more hopelessly tied up by IRC 6751(b) than it is now.  The decision not to look behind the signature, like the decision not to look behind the notice, provides a logical way for the Court to determine whether the requirements of the statute are met without having to get into the mind of the manager.

Petitioners here made logical arguments that the actions of the manager at a later stage indicated that he paid little or no attention to the document at the time he signed it.  Proving that, however, does them no good. 

Graev’s Long Shadow: Section 6751(b) and Supervisory Approval of Penalties

Today we welcome guest blogger Professor Monica Gianni. Professor Gianni serves as an Associate Professor in the Department of Accounting of the David Nazarian College of Business and Economics at California State University, Northridge. She is the successor author to Volume 6, Tax Practice and Procedure, of the Bittker & Lokken treatise on Federal Taxation of Income, Estate and Gifts and Of Counsel at Davis Wright Tremaine LLP. She wrote to me and asked if I could mention her article forthcoming in The Tax Lawyer – a publication of the ABA Tax Section.  I suggested that she might do a better job of describing her article than me and persuaded her to write a description herself.  She writes on the penalty litigation that has consumed the Court – and this blog – for the past few years.  Keith

As a reader of this blog, you have undoubtedly read numerous posts on Section 6751(b). Section 6751(b) requires supervisory approval in writing prior to assessment of certain penalties. Enacted in 1998 as part of the IRS Restructuring and Reform Act, the statute’s purpose was to prevent IRS agents from using penalties as bargaining chips. The section remained essentially dormant for over 20 years, with both the IRS and taxpayers accepting the position that approval needed to be obtained only prior to assessment. The trilogy of Graev cases and the decision of the Second Circuit Court of Appeals in Chai v. Commissioner changed the Section 6751(b) landscape completely, opening a Pandora’s box of taxpayers using Section 6751(b) to avoid penalties on the technicality of no-written-supervisory approval. Hundreds of court cases have followed, resulting in cases inconsistently interpreting Section 6751(b) and well-counseled taxpayers avoiding tax penalties.

I’ve written an article on this subject, which is due to be published in the next volume of The Tax Lawyer—Supervisory Approval of Penalties: The Opening of a Graev Pandora’s Box. The article tries to bring some order into the case law that has resulted from a badly drafted statute. (You can download the article here). After examining the current state of case law, the article concludes by recommending that the statute be repealed. Internal IRS procedures can address issues with the conduct of IRS employees while not opening the door to taxpayers using a technicality to avoid penalties and IRS employees potentially imposing penalties overbroadly in their attempts to comply with Section 6751(b). While others argue that repeal is not the answer, there seems to be agreement that something needs to be done. As Keith Fogg has pointed out, if the statute isn’t repealed, “maybe we will still be litigating Graev cases into the next decade helping to provide a never-ending source of blog posts.” 

read more…

The bulk of the litigation on this section has addressed when supervisory approval must be given to comply with Section 6751(b). The Tax Court has taken an expansive interpretation of the statute in favor of taxpayers, generally requiring that supervisory approval be obtained prior to the first formal communication to the taxpayer advising that a penalty will be imposed. The Circuit Courts of Appeals have started to disagree with the Tax Court. The Ninth Circuit, in Laidlaw’s Harley Davidson Sales, Inc. v. Commissioner, held that approval is required at the earlier of the assessment of the penalty or before the supervisor “loses discretion whether to approve the penalty assessment.” More recently, the Eleventh Circuit in Kroner v. Commissioner and Carter v. Commissioner reversed the Tax Court, holding that approval is required only before the assessment of penalties.

Rather than examine those decisions one more time, this post looks at procedural requirements of supervisory approval that have resulted from numerous Tax Court decisions in actions brought by taxpayers for penalty relief based on inadequate supervisory approval. First, what is required of a supervisor to fulfill the penalty-approval requirement? The simple answer is nothing but the approval itself. No cross-examination of the supervisor by the taxpayer is required, no reasonable-cause defense by the taxpayer has to be presented first, and there is no requirement that the “thought process” of the supervisor be analyzed or that her review of the penalty have been “meaningful.” The supervisor does not have to consider the merits of the penalty determination, does not have to have real estate expertise for a valuation penalty, and can even approve a valuation penalty before receiving the appraisal report. As summarized in Belair Woods, LLC v. Commissioner, the penalty approval form itself does not have to “demonstrate the depth or comprehensiveness of the supervisor’s review.”

The next question is—how is approval shown? The approval, by the express language of the statute, must be in writing. That being said, an actual signature is not required, and approval can be shown by an electronic signature or even by e-mail. If the approval form, however, shows no date of approval or the date is illegible, the taxpayer will prevail under Section 6751(b). The reason for the penalty on the approval form must be the same as contained in the Notice of Deficiency, and the specific penalty must be listed and not just a general statement that penalties are approved.

A further question is—who is the supervisor that must approve the penalty? Section 6751(b) requires that the taxpayer’s “immediate supervisor” approve the penalty, and this connection must be shown on the approval form. “Immediate supervisor” is not defined in the statute, and there are no regulations under this section. When faced with the issue, the Tax Court in Sand Investment Co. v. Commissioner determined that such supervisor “is most logically viewed as the person who supervises the agent’s substantive work on an examination, even if the examiner’s direct supervisor is someone else.” The IRS considers an acting supervisor to be the agent’s immediate supervisor if he has an approved Designation to Act or a Notification of Personnel Action on file.

If a taxpayer wants to challenge a penalty in court based on lack of IRS supervisory approval, are there any limitations? A taxpayer cannot raise the Section 6751(b) issue for the first time on appeal when the issue could have been raised in the Tax Court. Nor can the issue be raised for the first time at the district court level if it was not raised in administrative proceedings. For a TEFRA partnership action, Section 6751(b) must be raised at the partnership level and is not a partner-level defense. And, if a taxpayer enters into a closing agreement agreeing to the assessment of penalties rather than going to court, he waives any subsequent Section 6751(b) challenge.

The above describes just some of the procedural rules that have developed from numerous court cases post-Graev. Although there is more certainty now than there was prior to these cases, different results for taxpayers can occur depending on which circuit has venue over any ensuing appeal. Whether the statute has succeeded in preventing penalties from being used as bargaining chips seems to have become an irrelevant consideration, as taxpayers have used the statute to escape often well-deserved penalties.

2021 Year in Review – Graev

Maybe it’s too much to devote one year in review post to a single issue, but Graev has dominated case decisions the past few years and maybe, maybe not, is on its way out.  As we reported and blogged about here, the now stalled Build Back Better legislation has a provision that will eliminate Graev, not just going forward, but going back over 20 years.  Not since the retroactive elimination of the telephone excise tax has Congress tried to undo itself in such a grand way.  Since this may be the last hurrah for Graev, why not send it out in style or, if it remains, why not remind ourselves how a poorly worded piece of legislation can cause so much havoc.


Since the IRS has noticed the existence of IRC 6751(b), it seems now to have procedures in place to ensure that the immediate supervisor of the employee imposing the penalty actually approves the penalty imposition.  If the IRS has finally figured this out, why then repeal the legislation now and why does it receive a relatively high score from Congress for repealing it?  One suggestion concerns a whole bunch of old shelter cases that exist out there in pre-notice of deficiency status in which the IRS failed to follow the now more clearly defined rules of IRC 6751(b).  If true, it becomes easier to see why the administration would push for retroactive repeal and why certain groups of taxpayers would push back.  While we contemplate what might happen in the future, let’s look at the more important Graev decisions of 2021.

Graev and the Fraud Penalty

I posted on a decision that troubled me because when the IRS pursues a taxpayer criminally, the case goes through a myriad set of approvals.  Yet in Minemyer v. Commissioner, T.C. Memo 2020-99 – a case that took 10 years to decide – the Court found that the IRS did not follow IRC 6751(b) and stripped off the civil fraud penalty following a criminal tax case.  I wrote about the Minemyer case here and expressed surprise that IRC 6751(b) would stop the application of the fraud penalty in a case that involved a prior prosecution of the taxpayer, since the assertion of the fraud penalty following prosecution occurred automatically, with the hands of the agent and the agent’s supervisor essentially tied.  Of course, the statute does not specifically address prior criminal cases or create any special exception for them.

The Tax Court followed up the Minemyer case with a precedential decision in Beland v. Commissioner, 156 T.C. No. 5 (2021), where the Tax Court determined that the fraud penalty the IRS sought to assert failed the requirements of IRC 6751(b) allowing the taxpayers to avoid the 75% penalty proposed by the IRS without getting to the merits.  The Court issued this opinion granting partial summary judgment on the fraud issue five years after the case was filed. 

When a revenue agent seeks to impose the fraud penalty, the agent must send the case from exam over to obtain approval from the fraud technical advisor (FTA).  The FTA is a Small Business/Self Employed revenue agent specially trained on tax fraud issues. The IRS set up the system of having agents refer cases to FTAs so that an investigator trained specifically in fraud detection could determine if the revenue agent had gathered enough information to support the fraud penalty and to allow the FTA to determine if this case should chart a path toward criminal prosecution prior to imposition of the civil fraud penalty. See IRM (08-12-2016).

Requiring that the imposition of the fraud penalty first go through an FTA seems to provide even better protection against the use of the fraud penalty as a bargaining chip than having the immediate supervisor sign off on the penalty, but the statute has a specific structure applicable to all penalties.  Striking the fraud penalty in this situation may be part of what’s causing Congress to rethink its passage of IRC 6751(b), but for the reasons discussed in a post by Nina Olson, that seems too radical a fix to a problem that it could resolve with better statutory language.

Graev and the Early Withdrawal Excise Tax

Pulling money out of a retirement account before reaching 59 and ½ and without meeting one of the statutory exceptions in IRC 72(t) triggers a 10% excise tax usually referred to as a penalty and determined by bankruptcy courts to be a penalty for purposes of priority classification.  In Grajales v. Commissioner, 156 T.C. No. 3 (2021), the Tax Court determined that the 10% exaction imposed under IRC 72(t) is not a penalty for purposes of whether the IRS must obtain supervisory approval prior to its imposition.  The amount at issue in this precedential opinion was $90.86 and the case was litigated by Frank Agostino, the godfather of IRC 6751(b) litigation. See Frank’s brief here, and the government’s here.  Frank lost this one but given the way that most people look at this exaction, his arguments were not illogical.

Conservation Easement Cases

In Oconee Landing Property LLC et al v. Commissioner, Dk. No. 11814-19, the Tax Court entered a very substantive order granting partial summary judgment to the IRS on the issue of penalty approval.  If the Court still designated orders, I suspect it would have designated this one.

The taxpayer does not argue in this case that the IRS did not obtain the penalty approvals prior to the communication with it that the IRS had asserted a penalty.  Although the prior approval issue exists in most IRC 6751(b) cases, here the issue focuses on the form and manner of the approval, particularly as it relates to summary judgment.  It asserts that the penalty lead sheet in the file “does not identify Ms. Smithson’s [the immediate supervisor] role … or even a date of signature.”

In this case, the approval occurred through email rather than by a signing of the same paper by the agent and the immediate supervisor.  This type of approval has no doubt become quite common during the pandemic while many employees and managers have been working remotely.  It could also be common in situations where the employee and the manager work out of different offices.  Obtaining acceptance of this type of approval is important for the IRS.  One hurdle it has here and in many other cases involves proving that the person signing the approval is, in fact, the immediate supervisor of the employee imposing the penalty.

When is Supervisory Approval Necessary

In Walquist v. Commissioner, 152 T.C. No. 3 (2021), the issue focused on the IRS’s Automated Correspondence Exam (ACE) software. ACE automatically processes taxpayer returns. In many cases, ACE handles returns from receipt to closing with “minimal to no tax examiner involvement.” In Walquist, ACE processed the taxpayer’s 2014 tax return, assessed a §6662 penalty, and issued the notice of deficiency automatically and without any human interaction. The Tax Court found that because the penalty was determined mathematically by a computer software program without the involvement of a human IRS examiner, the penalty was “automatically calculated through electronic means.”

This decision creates a dichotomy between low-income taxpayers whose cases are regularly handled by somewhat automated processes and higher income taxpayers whose cases are not.  The Court did not need to issue a precedential opinion in a case in which the taxpayers were unrepresented tax protestors, yet decided to do so despite the inability of the adversarial process to work effectively.  The Court heard only from the government and clearly expressed displeasure at the unfounded arguments advanced by these taxpayers.  The decision leaves a bad taste in my mouth for the way it casually treats an issue involving many low-income taxpayers without giving lawyers for low-income taxpayers the opportunity to present arguments explaining why this result should not attach.  I have a working paper on the topic of precedential opinions in pro se cases and possible solutions to the creation of precedent where only the government has a real voice.

Congress to Consign IRC 6751(b) to the Graev?

Carl Smith brought to my attention that one of the provisions in the tax bill currently working its way through Congress proposes to repeal IRC 6751(b) and he provided me with the title to this post.  I joked with others on the email thread that such a repeal could spell the end of our blog.  We have written so many posts on this poorly drafted law that having it repealed could significantly reduce the topics upon which we would write.

The other thought that went through my head was that the IRS had regained its voice in influencing tax legislation.  In the 1960s, 1970s, and into the 1980s perhaps up to the first Taxpayer Bill of Rights legislation, the IRS had less to fear from adverse judicial precedent because it could generally fix bad precedent by going to Congress to reverse the problem caused by adverse precedent.  This influence seemed particularly present when Wilbur Mills chaired the Ways and Means Committee.  I have wondered about the impact of his dip in the Tidal Basin with Fanne Foxe on tax legislation and its trajectory.  For the past couple decades, or more, the IRS seemed to have lost that ability, although there are some notable exceptions.  Maybe it is returning.  Congressman Neal, be careful where and with whom you go swimming.


Congress passed 6751(b) as a part of the Restructuring and Reform Act of 1998 (RRA 98) in an effort to curb perceived abuse of penalty provisions to pressure taxpayers into conceding tax liabilities.  No doubt some IRS agents used penalties or the prospect of penalties to cajole taxpayers into settlements.  I never perceived this as a common practice, but I may not have been in the best position to observe behaviors regarding this issue.  Whether real or mostly imagined, the statute drafted to fix the alleged problem was drafted by someone with little or no knowledge of tax procedure.  It caused the Tax Court to twist itself into knots in order to interpret it and led Judge Holmes in dissenting from the interpretation in the Graev case to predict that the decision would led to many unpredicted and poor outcomes that he labeled Chai ghouls after the Second Circuit decision on this issue.  He was right, but that’s not to say the majority of the court got the Graev case wrong.  The statute is so poorly worded that no judge could draw the precise meaning.

One thing that saved the tax procedure world from 6751(b) for most of its life was that everyone ignored it.  Neither the IRS nor the bar seemed to pay attention to its requirements, whatever they were or are, for over 15 years.  Thanks to Nina Olson and the annual report to Congress and to Frank Agostino, litigation finally began to seek to provide meaning to the statute.  The cases have led to results that don’t necessarily follow the goal of the statute and that let many taxpayers off the hook for penalties, not because the IRS used them inappropriately as bargaining tools but because it failed to secure the appropriate approvals.  In some instances, the IRS would have had difficulty knowing what approvals would have been appropriate at the time of the imposition of the penalty since the court interpretation of the statute came several years thereafter.  Of course, the IRS could have headed this off with timely regulations that sought to interpret the provision and set up clear rules to follow but it did not create such regulations and suffered in several cases for its inaction.

Maybe this obit for 6751(b) comes too early.  Proposed legislation does not change the law.  Here is the language of the proposed repeal which contains its own surprise:


(a) REPEAL OF APPROVAL REQUIREMENT.—Section 6751, as amended by the preceding provision of this Act, is amended by striking subsection (b).  

(b) QUARTERLY CERTIFICATIONS OF COMPLIANCE WITH PROCEDURAL REQUIREMENTS.—Section 6751, as amended by subsection (a) of this section, is amended by inserting after subsection (a) the following new subsection:  

‘‘(b) QUARTERLY CERTIFICATIONS OF COMPLIANCE.—Each appropriate supervisor of employees of the Internal Revenue Service shall certify quarterly by letter to the Commissioner of Internal Revenue whether or not the requirements of subsection (a) have been met with respect to notices of penalty issued by such employees.’’.  



The amendment made by subsection (a) shall take effect as if included in section 3306 of the Internal Revenue Service Restructuring and Reform Act of 1998.  


The amendment made by subsection (b) shall apply to notices of penalty issued after the date of the enactment of this Act.

Notice in subsection (c)(1) of the proposed legislation that it repeals 6751(b) back to 1998.  Wow!  While wiping this statute off the books makes Congress look better and gets rid of a provision that makes it look incompetent, few examples exist of such a period of retroactivity.  Might the repeal itself, assuming it occurs, create yet more blog post opportunities for Procedurally Taxing?  Maybe the repeal is a good thing for the blog since the number of different permutations of 6751(b) may be dwindling, reducing the number of future posts if the current law continues unchanged.  A repeal such as this could be just the ticket for an additional round of posts as taxpayers whose cases are currently pending at some point on the continuum of audit or litigation will want to fight for the right to have their penalty removed.

Note also that section 6751(a), which provides that “The Secretary shall include with each notice of penalty under this title information with respect to the name of the penalty, the section of this title under which the penalty is imposed, and a computation of the penalty,” would not be repealed by the proposed legislation. The legislation would merely replace subsection (b) with a certification requirement within the IRS by managers on a quarterly basis as to their employees’ general compliance with subsection (a) – something likely of no use to taxpayers.

What about the sentiment that caused the essentially unanimous passage of RRA 98 regarding using penalties as an inappropriate bargaining chip?  Has the IRS cleaned up its act in this regard over the past quarter century?

We almost never write about pending legislation but with the opportunity to use the title to today’s post coupled with the possible loss of one of the most productive post producers, it seemed appropriate in this instance.  Now that you know about the proposal, you can start your lobbying efforts on behalf of the legislation or against it.  No matter which way this plays out, 6751(b) has to remain among the top few provisions as the worst ever drafted in the tax procedure realm.

Graev Stops Assertion of the Fraud Penalty

The whole month of February passed without us writing a blog post on another Graev issue.  While we did not write a Graev based post in February 2021, the Tax Court issued another precedential opinion on the issue of supervisory approval.  In Beland v. Commissioner, 156 T.C. No. 5 (2021) the Tax Court determined that the fraud penalty the IRS sought to assert failed the requirements of IRC 6751(b) allowing the taxpayers to avoid the 75% penalty proposed by the IRS without getting to the merits.  The Court issued this opinion granting partial summary judgment on the fraud issue only five years after the case was filed.  Presumably, it will now hold a trial at some point in the future to determine what amount of tax the taxpayers owe.

The Tax Court had previously ruled in a non-precedential opinion that the IRS flubbed the approval process in a fraud penalty case in Minemyer v. Commissioner, T.C. Memo 2020-99 – a case that took 10 years to decide.  I wrote about the Minemyer case here and expressed surprise that IRC 6751(b) would stop the application of the fraud penalty in a case that involved a prior prosecution of the taxpayer, since the assertion of the fraud penalty following prosecution occurred automatically, with the hands of the agent and the agent’s supervisor essentially tied.  Of course, the statute does not specifically address prior criminal cases or create any special exception for them.

Beland does not involve the imposition of the fraud penalty in a prior criminal case but does bring out in its discussion the special processing that occurs when the IRS seeks to impose the fraud penalty.  As I will discuss below, that processing generally provides the protections sought by the legislation but does not follow the format adopted by the statute. For another take on the decision read Bryan Camp’s post on this case as part of his lessons from the Tax Court.  As usual Bryan has great insight on the case.


The opinion does not really discuss what Mr. Beland did to cause the IRS to impose the fraud penalty as the issue here arises on a motion for summary judgement concerning the timely approval of the penalty.  When a revenue agent seeks to impose the fraud penalty, the agent must send the case from exam over to obtain approval from the fraud technical advisor (FTA).  The FTA is a Small Business/Self Employed revenue agent specially trained on tax fraud issues. The IRS set up the system of having agents refer cases to FTAs so that an investigator trained specifically in fraud detection could determine if the revenue agent had gathered enough information to support the fraud penalty and to allow the FTA to determine if this case should chart a path toward criminal prosecution prior to imposition of the civil fraud penalty. See IRM (08-12-2016).

Requiring that the imposition of the fraud penalty first go through an FTA seems to provide even better protection against the use of the fraud penalty as a bargaining chip than having the immediate supervisor sign off of the penalty, but the statute has a specific structure applicable to all penalties.  It does not recognize the special handling the IRS has long pursued when it perceives that fraud exists.  In addition to sending the case through the FTA, before the agent can issue a notice of deficiency, the agent must obtain prior approval from Chief Counsel’s office. See IRM (04-29-2016).  This approval occurs because the burden of proof in fraud cases rests on the IRS and before sending out a notice of deficiency, the IRS wants to make sure that it has the necessary proof to sustain the fraud penalty.

The opinion notes that the agent did seek assistance from the FTA.  It even says that the agent and her supervisor referred the case to the FTA.  If the supervisor signed off on the referral to the FTA, an argument exists that the supervisor has engaged in at least some form of approval of the penalty given the reason for the referral to the FTA.

Here, the FTA requested that the agent summons petitioners to appear before them.  I found this an odd request.  I do not remember seeing a summons in this situation when I worked for Chief Counsel but perhaps procedures have changed.  After some delay due to the birth of a child, petitioners came to the IRS for a meeting which not only served as the response to the summons but also as the closing conference.  As is normal, the agent attempted to obtain the agreement of the petitioners to the proposed adjustments.  And as is normal in a fraud case, the petitioners declined to sign up for the tax and fraud penalty.  The number of taxpayers who willingly consent to the assessment of the fraud penalty cannot be a very high number.

After the meeting the agent obtained the signature of his immediate supervisor on the Civil Penalty Approval Form.  Keep in mind that this all occurred in 2015 before the Graev opinion and the attention on IRC 6751.  The Tax Court had not yet, and would not for several more years, opine on when an initial determination occurred.  It’s possible to fault the IRS for not getting in front of the IRC 6751 issue shortly after it because law in 1998, but the agent obtained the supervisory signature in a reasonable time frame and had already, presumably, obtained approval to assert fraud from the FTA.  Nothing in the agent’s actions appears to point toward use of the penalty as a bargaining chip.

Between the time of the supervisory signature and the hearing on the motion for summary judgment, the Tax Court held in Clay v. Commissioner, 152 T.C. at 249 that an initial determination occurred when the IRS issued an agent’s report proposed the penalties and included the 30 day letter.  In Oropeza II, 155 T.C. at (slip opinion pp. 4-6) the Tax Court held that the agent’s report (together with the letter saying the taxpayer would not be offered a trip to Appeals due to the running of the statute of limitations) served as the initial determination.

Here, the Court finds that presenting petitioners with the agent’s report at the closing conference “sufficiently denoted a consequential moment in which RA Raymond had made the initial determination to impose the fraud penalty.”  The Court went on to explain:

While the revenue agents in Clay, Belair Woods, and Oropeza II sent the taxpayers the RARs through the mail, we have never held that an initial penalty determination must be communicated by letter. Rather, the Court’s focus is on the document and the events surrounding its delivery that formally communicate to the taxpayer the IRS’ decision to definitively assert penalties.

The Court also noted that using the summons to essentially force the petitioner to attend the closing conference added to the formality of the meeting.  So, the Court pushes the time period of the initial determination back before the formal issuance of the 30-day letter.  This may open up additional opportunities for taxpayers seeking to attack the imposition of a penalty.

Although the Court knew of the referral to the FTA, it does not talk about that process in reaching its conclusion.  In discussing the arguments made by the IRS, it does not appear that the IRS sought to argue the referral process used in fraud cases played in role in determining when approval occurred.

This decision is important for taxpayers in cases where the immediate supervisor has signed prior to delivery of the agent’s report, but the agent has had some formal meeting before the delivery of the report.  That aspect of the decision makes sense to me, given the purpose of the statute in preventing the IRS from using penalties as a bargaining chip.  Where I have trouble in the cases involving the fraud penalty is the countervailing policy argument that the IRS has taken other steps to protect the taxpayer from having the fraud penalty used as a bargaining chip.

Graev and the Early Withdrawal Exaction under IRC 72(t)Graev and the Early Withdrawal Exaction under IRC 72(t)

In Grajales v. Commissioner, 156 T.C. No. 3 (2021) the Tax Court determined that the 10% exaction imposed under IRC 72(t) that most people colloquially call a penalty is not a penalty for purposes of whether the IRS must obtain supervisory approval prior to its imposition.  The taxpayer in this precedential opinion involving $90.86 was represented by Frank Agostino who pioneered the use of IRC 6751(b)

The case pits the characterization of IRC 72(t) in Tax Court cases against its characterization in bankruptcy cases.  In many ways it presents the mirror image of IRC 6672 which goes under the name of trust fund recovery penalty and which the Tax Court treats as a penalty, but bankruptcy law treats as a tax.  These are not the only two code sections where the label as tax or penalty depends on the context and where that context has applications that can result in significant differences based on the label.


The Tax Court has consistently treated IRC 72(t) as imposing a tax.  The issue has come up in several contexts prior to the challenge under IRC 6751(b).  The decision in Grajales continues the Tax Court’s consistent treatment of the provision.  The court notes:

In contexts apart from the application of section 6751(b)(1), this Court has held repeatedly that the section 72(t) exaction is a “tax” and not a “penalty”, “addition to tax”, or “additional amount”. See, e.g., Williams v. Commissioner, 151 T.C. 1, 4 (2018) (holding that the section 72(t) exaction is not a “penalty, addition to tax, or additional amount” within the meaning of section 7491(c) for purposes of placing the burden of production); El v. Commissioner, 144 T.C. 140, 148 (2015) (same); Dasent v. Commissioner, T.C. Memo. 2018-202, at *7 (same); Summers v. Commissioner, T.C. Memo. 2017-125, at *5 (same); Thompson v. Commissioner, T.C. Memo. 1996-266, 1996 WL 310359, at *7 (holding that the section 72(t) exaction is a “tax” rather than a “penalty” for purposes of the joint and several liability provision of section 6013(d)(3)); Ross v. Commissioner, T.C. Memo. 1995-599, 1995 WL 750120, at *6 (same).

With that history of the treatment of IRC 72(t) in Tax Court cases, the opinion provides no surprises, as it methodically works through the reasoning for finding that the provision should receive treatment as a tax.  Looking at the treatment of the provision in other contexts within the tax code, the Court finds consistency in the description of this provision:

First, section 72(t) calls the exaction that it imposes a “tax” and not a “penalty”, “addition to tax”, or “additional amount”. Second, several provisions in the Code expressly refer to the additional tax under section 72(t) using the unmodified term “tax”. See secs. 26(b)(2), 401(k)(8)(D), (m)(7)(A), 414(w)(1)(B), 877A(g)(6). Third, section 72(t) is in subtitle A, chapter 1 of the Code. Subtitle A bears the descriptive title “Income Taxes”, and chapter 1 bears the descriptive title “Normal Taxes and Surtaxes”. Chapter 1 provides for several income taxes, and additional income taxes are provided for elsewhere in subtitle A. By contrast, most penalties and additions to tax are in subtitle F, chapter 68 of the Code.

After establishing why the Court should treat 72(t) as a tax and therefore not impose on the IRS a requirement that it obtain supervisory approval prior to its imposition, the Court addresses the arguments presented by petitioner.

It first rejects petitioner’s argument that it should change its practice of following the label given to the liability in the tax code.  It then rejects petitioner’s argument that the Supreme Court’s decision in Nat’l Fed’n of Indep. Bus. v. Sebelius (NFIB), 567 U.S. 519 (2012) in which the Supreme Court recharacterized the label given to a liability imposed by the Affordable Care Act in order to determine that the Anti-Injunction Act did not bar it from making a decision.  Finally, it addressed the numerous bankruptcy court decision holding that 72(t) imposes a penalty for bankruptcy purposes.

The Tax Court does not reject the characterization of 72(t) as a penalty for bankruptcy purposes any more than it embraces the treatment of the TFRP as a tax.  Back in 1979 the Supreme Court first characterized the TFRP as a tax for purposes of bankruptcy.  The decision in United States v. Sotelo, 436 U.S. 268 (1978) held that for bankruptcy purposes TFRP was a tax which has significant implications in the payment of that liability through bankruptcy.  In Chadwick v. Commissioner, 154 T.C. No. 5 (2020), blogged here, the Tax Court determined that the IRS must follow the requirements of IRC 6751(b) in TFRP cases.

Not only is the decision in Grajales consistent with prior Tax Court decisions regarding 72(t), it is consistent with the way the Tax Court and the courts interpreting the bankruptcy laws have treated tax provisions labeled as tax or labeled as penalties.  The Tax Court consistently treats these cases as following their designations in the Tax Code.  Courts interpreting the same provisions for bankruptcy purposes have consistently looked behind the label on the provision to the effect or function of the provision and treated provisions labeled as tax as though they were penalties and vice versa where the circumstances supported a different label.

I am not troubled by the Tax Court following the form while courts interpreting the bankruptcy code follow function.  The systems serve different purposes.  As long as each applies the rules consistently in their realm, taxpayers and practitioners can adjust to the realities of the situation in which they find themselves.  The prior approval process required by 6751(b) lives in the tax code and applies to tax provisions labeled as penalties.  The tests developed in the bankruptcy code to treat certain provisions in a manner differently than the label they carry serves a different purpose. 

In Grajales the Tax Court addressed a provision most people call a penalty but Congress did not.  The opinion logically follows the path the Tax Court has followed in the path and it is a path that does not need to be changed.  At the same time the petitioner’s arguments here also logically pointed out the difference in treatment of tax provisions in other settings.  If we want consistency, it raises larger questions.

Tax Shelter Investor Relieved of Penalties by Graev

In the latest precedential opinion from the Tax Court on Graev, Oropeza v. Commissioner, 155 T.C. No. 9 (2020), the Court refuses to impose any penalty on the petitioner despite his investment in a tax shelter.  The case receives precedential treatment because it finds yet another way in which the Graev case impacts the IRS as it goes to impose penalties.  Mr. Oropeza, as have several beneficiaries of the Graev precedent, benefits from the age of this case.  The tax year is 2011 and he docketed his Tax Court petition in 2015.  At the time the revenue agent worked this case, Graev had not burned into the IRS consciousness the need to obtain penalty approval prior to sending out the revenue agent’s report.  The statute came into existence in 1998.  So, it’s hard to build up significant sympathy for the IRS regarding the way it handled penalties after Congress had spoken, but timing here does matter.

If you are interested in this case, you should also read the excellent post by Bryan Camp in his Lessons from the Tax Court series.  The IRS has also issued some guidance related to Graev that should be noted.  I was alerted by Jack Townsend of a new IRM provision that provides important guidance to revenue agents in this area: (10-19-2020)

Timing of Supervisory Approval

For all penalties subject to IRC 6751(b)(1), written supervisory approval required under IRC 6751(b)(1) must be obtained prior to issuing any written communication of penalties to a taxpayer that offers the taxpayer an opportunity to:

  • Sign an agreement, or
  • Consent to assessment or proposal of the penalty

Jack has blogged on this provision and you can find his post here.


Mr. Oropeza invested in microcaptive insurance.  The upcoming Supreme Court argument in CIC Services presents the same type investment though, as with Mr. Oropeza, the case does not directly involve the taxation of the investment itself.  The revenue agent audited Mr. Oropeza’s return and proposed, in Letter 5153 and a revenue agent report dated January 14, 2015 accuracy related penalties under 6662(a) of 20%.  Prior to the sending of this letter, the immediate supervisor of the revenue agent had not signed the approval of the penalties and did not sign it until two weeks later.

On May 1, 2015, the revenue agent changed his recommendation proposing a penalty under 6662(i) for 40% for a transaction lacking economic substance.  The revenue agent’s immediate supervisor signed the form approving this increase.  A notice of deficiency issued on May 6 focused on the 40% penalty with the 20% penalty as a fallback position. 

Based on Belair Woods, LLC v. Commissioner, 154 T.C. 1, 14-15 (2020) interpreting IRC 6751(b), we know that the 20% penalty will fail because the supervisory approval came after the agent transmitted the IRS position to the taxpayer.  At the time it filed its original brief in the Oropeza case, the government attorneys did not have the benefit of the Belair Woods decision and argued that the notice of deficiency was the appropriate date for determining whether approval of the immediate supervisory had occurred.  The IRS argued that it met the requirement of the statute.  In subsequent briefing on the case, it had the opportunity to address the Belair Woods decision.

At issue here is the impact of 6751(b) on the 40% penalty for which the revenue agent obtained approval prior to notifying the taxpayer.  (Some uncertainty on this issue exists because the immediate supervisor failed to date the approval form as he signed it; however, his testimony supported the fact that he signed on the same day as the revenue agent.  Of course, the date of his signature does not matter given the outcome of the case.)  The IRS argued that imposition of the penalty under IRC 6662(i) represented a separate penalty properly approved before notifying the taxpayer.  The Tax Court decided the case on cross motions for summary judgment.

The IRS argued that Oropeza differed from Belair because the letters offering a conference with Appeals differed slightly.  The Tax Court rejected the attempt to make a distinction based on the type of letter offering an Appeals conference.  It stated:

The Letter 5153 clearly communicated the same message to petitioner: It told him that he could now go to Appeals, but only if he first executed a Form 872 that would give Appeals enough time to consider his case.

So, the IRS communicated the initial determination to the taxpayer in the Letter 5153 and sent it prior to supervisory approval.  Does that lock the IRS into failure on the penalty issue or does the change in penalty position and the obtaining of a signature prior to communication with the taxpayer of the new penalty provide the IRS with a reprieve?

In approaching this issue, the Tax Court stated:

Litigants sometimes refer to section 6662(i) — like section 6662(h), applicable in the case of a “gross valuation misstatement” — as imposing a “40% penalty.” As the statute’s text makes clear, however, section 6662(i) does not impose a distinct penalty. It simply increases the rate of the penalty imposed by section 6662(a) and (b)(6) for engaging in a transaction lacking economic substance.

Given the statutory structure, we think the proper analysis requires that we answer two questions. The first is whether the RAR should be read as asserting a penalty under section 6662(a) and (b)(6) for engaging in a transaction lacking economic substance. If so, that penalty did not receive timely approval. The second question is whether, if the base-level penalty was not timely approved, the IRS can satisfy section 6751(b)(1) by later determining that section 6662(i) applies because the transaction was not disclosed on the return.

The Tax Court examined the IRS communication regarding the 20% penalty, noting that a taxpayer receiving the correspondence from the IRS would struggle to know which of the grounds of IRC 6662 the IRS sought to apply including “a transaction lacking economic substance.”  The IRS argued that while the revenue agent’s report might have been ambiguous, additional evidence shows what it meant.  The Court declines to look behind the information provided to the taxpayer and that information included the possibility that the penalty related to a transaction lacking economic substance.

The Tax Court then looked at IRC 6662(i).  It concludes that this subsection does not impose a distinct penalty but simply increases the rate of the penalties imposed under 6662(a) and (b)(6).  Based on this analysis, the Court finds that 6662(i) serves as something similar to an aggravating factor in criminal law leading to a higher punishment.  It moves from that conclusion to the conclusion that the initial penalty proposed in Letter 5153 operates as the touch point that it must review to determine if the IRS timely approved the penalty.  Looking at the transaction at that point, the IRS fails.

Linking the penalties imposed under 6662 in this way may benefit the IRS if it timely approves the first notice of a 6662 penalty but does not obtain a timely approval of an increase with the categories of 6662 penalties.  Maybe the next case will fall in the IRS’s favor because of this linkage.  There may be other implications of the linkage determined in Oropeza as we look at penalty cases in the future.  Professor Camp’s post does a good job of looking at possibilities and inconsistencies.  This case may have implications that reach beyond mere penalty approval.

Splitting More Graev Hairs – Tax Shelter Generic Settlement Offers

In yet another precedential opinion involving the interpretation of IRC 6751(b) Tax Court Judge Greaves decides a Graev issue in favor of the IRS.  In Thompson v. Commissioner, 155 T.C. No. 5 (2020) the issue turns on the impact of a settlement offer made to the taxpayer prior to supervisory approval of the penalty proposed by the IRS.  The Tax Court determines that proposing to settle a penalty does not equal an initial determination under the statute.  Because the settlement proposal did not equal an initial determination, the IRS does not lose the penalty for making the settlement offer before obtaining permission of the revenue agent’s immediate supervisor.


 The taxpayer invested in a distressed asset trust transaction (DAT transaction) which the IRS had targeted for special treatment as an abusive transaction.  For these types of targeted transactions, the IRS sometimes develops special settlement initiatives to try to run multiple cases through the same procedure, avoiding as much time and trouble as possible.  A revenue agent sent the taxpayer a letter at the beginning of the examination laying out the terms of the settlement, including the amount of the penalty the taxpayer would need to pay for the IRS to accept a deal.  The court described the letter as the revenue agent stating he was:  “aware that * * * [petitioners] participated in an abusive transaction” and offered them “the opportunity to resolve * * * [their] tax liabilities associated with that transaction in accordance with the terms set forth in * * * [Announcement 2005-80 (Oct. 28, 2005)].”

The court further stated that this initial “letter did not identify a tax period or tax form to which it related, provide an underpayment amount, or request petitioners’ consent to assessment and collection. Petitioners did not accept the settlement offer in the 2007 letter.”

The Revenue Agent mailed a second letter 20 months later in 2009 when the taxpayers did not respond to the initial letter.  Similar to the first letter, the 2009 letter did not identify a tax period, form number, or underpayment amount and did not request petitioners’ consent to assessment and collection. The letter explained that if petitioners did not accept the settlement terms, the IRS would “complete its examination and fully develop the facts” of petitioners’ case and “impose applicable penalties under the Internal Revenue Code.”

Both letters went out prior to the determination by the agent’s immediate supervisor regarding the penalty as part of a larger determination by the IRS on how to treat penalties in this situation.  The agent and the manager’s hands were bound by the overall agency decision on how to treat these cases.  (In that regard this case has some similarities to the recent decision in Minemyer v. Commissioner that ended up with a different result.)

Because the taxpayers did not respond to the second letter, the revenue agent proceeded to examine their return.  In doing so he determined that they owed tax and both an accuracy related penalty under IRC 6662(h) as well as a penalty under 6662A.  He obtained the approval of his immediate supervisor who was acting in the position in December 2009.  A notice of deficiency was sent in 2012.  This case has been in the Tax Court since March of 2013.  Here is a link to the docket sheet if you want to follow along the slow path this case has taken to decision for liabilities on the years 2003 to 2007. 

The court describes three arguments raised by the Thompsons.  First, they argued that the settlement letter served as an initial determination and required the approval of the revenue agent’s supervisor, since it raised the issue of penalties.  Second, they argued that the approval was signed by an acting immediate supervisor who would not have provided meaningful review.  Third, they argued that to the extent of any ambiguity, because this is a penalty situation, the ambiguity should be resolved in their favor.  The court addressed and rejected each argument. 

With respect to the need for the IRS to obtain supervisory approval prior to sending out the settlement letter, the court provided a very reasoned basis for explaining why the statute did not require the IRS to do so prior to sending this type of letter:

The offer letters in this case do not reflect an “initial determination” because they do not notify petitioners that Exam had completed its work. Rather than determining that petitioners are liable for penalties of specific dollar amounts, subject to review by Appeals or the Tax Court, each letter offers to settle penalties arising from the DAT transaction on certain terms, including substatutory penalty rates, which are based not on an audit but on Announcement 2005-80. When petitioners failed to accept the offers, RA Damasiewicz still had work to do — the 2009 letter explicitly says the IRS had not completed its examination or fully developed the facts of petitioners’ case. Furthermore, the 2009 letter warns that declining the settlement offer would result in “applicable penalties,” without stating which penalties, if any, might end up being “applicable” to petitioners’ facts. An offer letter like the ones at issue does not require supervisory approval because it is not a “determination” at all, but a preliminary proposal of the revenue agent within an ongoing examination.

The court’s reasoning makes good sense to me.  The opposite result would create quite a shakeup at the IRS in the way it processes tax shelter settlement offers.  Just because it would shake up the long-standing practice does not make the practice one that requires preserving, but the letters sent in cases such as this do not tell the taxpayer exactly what the IRS has decided in their case, other than that their case is one the IRS has identified as a tax shelter.  The taxpayers probably already knew that when they invested in the shelter and signed numerous documents acknowledging the risk.  The initial letter lets the taxpayer know the risk has increased significantly but provides few specifics. 

The second issue regarding the significance of having the supervisory approval form signed by an acting supervisory rather than a permanent supervisor gets very little comment from the court.  If the IRS had to get penalty forms signed by permanent supervisors rather than actors, probably half of the agents could not move forward with their cases.  Acting supervisors are everywhere.  While they may or may not offer the same level of review as permanent supervisors, the statute does not prohibit acting supervisors from being the ones to approve the penalty form.  The court’s short treatment of this argument makes sense to me.

The third issue concerns the rule of lenity and the court cites the Rand case where Andy Roberson argued for the application of this rule.  The court finds that the situation in this case does not operate as one where the rule of lenity can stop the application of the penalty.  The decision provides no surprises in this regard.  Tax shelter investors will have a hard time invoking this rule.  The level of sympathy they evoke does not equal that of the low income taxpayers involved in the Rand case, who saw the IRS impose a penalty upon them for claiming the earned income tax credit in a way they could not support.

The Thompson case adds another interpretation to the Graev situation and receives precedential status as a result.  We now know that general letters offering a broad settlement initiative that includes penalties do not require the type of supervisory approval described in 6751(b).  I understand why petitioners would want to make this argument.  The case presents a situation like the Minemyer case, where hanging around in the Tax Court for long enough can allow a taxpayer to benefit from creative arguments that come out long after the filing of the petition.  Here, the taxpayers do not benefit from the creative argument but testing out the possibility certainly made sense.