To the uninitiated, §6662 seems to impose a single accuracy-related penalty that varies in amount depending on just how badly the taxpayer screwed up. For example, here’s the Wikipedia description: “This penalty of 20% or 40% of the increase in tax is due in the case of substantial understatement of tax, substantial valuation misstatements, transfer pricing adjustments, or negligence or disregard of rules or regulations.” Note the singular “this penalty.”
Today’s post will initiate you. In Roderick M. Campbell and C. Sandra Campbell v. Commissioner, T.C. Memo. 2020-41 (April 7, 2020) (Judge Ashford) an IRS Supervisor’s approval of a accuracy-related penalty was insufficient to comply with §6751 because the approval form did not specify whether the approval was for the 20% or 40% penalty amount. Despite the technical win here, however, this may not be a strong case for taxpayers. Details below the fold.
Section 6751 is a poorly written statute that has been a drain on taxpayer, IRS, and judicial resources. I suppose some might see an upside: some of those resources have drained into the pockets of tax litigators. I give a short history of the statute and its problems in Lesson From The Tax Court: A Practical Interpretation of The Penalty Approval Statute §6751 (January 18, 2020). For today, you just need to know that §6751 requires that any IRS employee’s decision to impose a penalty be approved by that employee’s supervisor at some point before the IRS sends the taxpayer its official determination to impose the penalty. In an audit that generally means before the 30-day letter (if there is one) or 90-day letter. See Clay v. Commissioner, 152 T.C. 223 (2019).
Facts and Lesson
In Campbell the IRS audited Mr. and Ms. Campbell’s 2008 return, and sent them an NOD that asserted both a $286,000 deficiency and a §6662 penalty. The proposed penalty amount was calculated as 20% for that part of the deficiency attributable to Schedule E errors and 40% for that part of the deficiency attributable to Schedule A errors. Op. at note 3. That was apparently what the NOD said. Op. at 14. What the NOD did not say, however, was that the Revenue Agent’s boss had signed off on a Civil Penalty Approval form that contained only this language: “Penalty will be asserted with the 2008 Year.” Id.
When the taxpayers saw the penalty approval form they saw their chance at a windfall and they asked the Tax Court to dismiss the proposed penalties because the IRS had not complied with §6751.
Judge Ashford obliged. She noted that even though, to the untrained eye, the statute reads like it is single penalty, the eyes of the IRS employees should now have been trained by relevant judicial precedent to see that “the section 6662(a) and (h) penalties are distinct...and the initial determination under each subsection must be separately approved for purposes of section 6751(b)(1).” Judge Ashford cites Palmolive Bldg. Investors, LLC v Commissioner, 152 T.C. 75, 87 (2019) as the go-to case for that proposition.
The IRS failure here was that the Penalty Approval Form did not specify whether the agent was proposing the 6662(a) penalty or the 6662(h) penalty or both. Judge Ashford says “if written supervisory approval of an unspecified penalty was sufficient[,] examining agents would be free to assert any type of penalty after written supervisory approval was given, an action that section 6751(b)(1) was designed to prevent.” So splat! went the proposed §6662 penalties.
Three Comments: Of Bugs and Windshields
First, don’t get too happy about this result. To see why, let’s compare this case to Palmolive. At first blush they look very similar yet the taxpayers here won and the Palmolive taxpayers lost. Palmolive involved a partnership audit, so the relevant document was a FPAA and not an NOD. In preparing the FPAA the Revenue Agent proposed a §6662(h) penalty of 40% and a §6662(a) penalty of 20%. It is true the Palmolive penalties were proposed in the alternative whereas in Campbell they were both proposed to be applied to two different sources of the understatement. But the fundamental point is that in Palmolive, like here, the form signed by the supervisor did not specify which penalty was being approved. The Palmolive Court specifically noted that “the pages that bear the supervisors' signatures do not mention the specific penalties they approved.” 152 T.C. at 86. That seems very similar to the facts in today’s case.
The same generality that was fatal to the IRS here, however, was not fatal in Palmolive. The reason appears to be that the Palmolive Court was willing to link the signature to an attachment to the form and not just to the form. Wrote Judge Gustafson: “the Form 5701 that [the supervisor] signed did state "Estimated Flowthrough Penalty (See F886A)", [sic] and to that form were attached Forms 886A that do justify the gross overvaluation [6662(h) penalty] and negligence penalties [6662(a) penalty].” Id. at 86.
So in today’s case, if the IRS here had been able to show Judge Ashford that there were documents linked to the Penalty Approval Form that gave specificity to the generic language in the Form, I think the IRS would have won. I wonder why the IRS did not argue to Judge Ashford here that the NOD showing the proposed penalty calculations against the Campbells was just like the Form 886A in Palmolive and could serve the same function: giving specific content to the approval. Perhaps the NOD was not attached to the penalty approval form here? Judge Ashford appears to believe that the one-page form was all that the supervisor had before him. She notes that “the record does not include any other evidence of penalty approval for 2008.”
So this is likely a failure of proof, or else the approval procedure used in this case looks more like a bug than a feature.
Second, speaking of bugs, while the taxpayers here got a windfall, usually taxpayers will be the bug and not the windshield. You see that in Roth v. Commissioner, 922 F.3d 1126 (10th Cir. 2019), where the 10th Circuit Court of Appeals affirmed the Tax Court's opinion and adopted the Tax Court’s reading of §6662 as imposing two separate penalties for §6751 purposes. In Roth, the Revenue Agent proposed the (h) 40% penalty, received approval and sent out in a 30-day letter. The taxpayers then went to Appeals in non-docketed status. Eventually Appeals issued the NOD proposing only the §6662(a) 20% penalty (also properly approved). When the case went into Tax Court, the IRS Office of Chief Counsel attorney reviewed the case and, with proper supervisory approval, re-asserted the 40% penalty in the government’s Answer. Naturally, that meant the government would bear the burden of proof on that issue since it was not in the NOD but was raised by the government’s pleading. Tax Court Rule 142.
The Tax Court opinion explained why the NOD could not lock the IRS into a single penalty determination. Different functions in the IRS might make different determinations as to an appropriate penalty. To say the IRS could not propose a different penalty before the Tax Court would run counter not only to decades of practice but also to the Tax Court’s statutory grant of jurisdiction in §6214, which says that the Tax Court’s jurisdiction includes the power “to determine whether any additional amount, or any addition to the tax should be assessed, if claim therefor is asserted by the Secretary at or before the hearing or a rehearing.”
The 10th Circuit agreed and got there by recognizing that §6662 contains two penalties for purposes of the §6751 supervisory approval. Here's the meat of it:
“Even if we...conclude that the notice of deficiency constituted an "initial determination" that a 20% penalty under § 6662(a) should apply, it is not clear why the IRS...could not subsequently make another "initial determination" (with written, supervisory approval) that a different, 40% penalty should apply under § 6662(h). Each of these determinations would appear to be "initial" with respect to the actual penalty sought, whether under § 6662(a) or § 6662(h).” 922 F.3d at 1135.
In this, the 10th Circuit agreed with the Tax Court’s pronouncement in Palmolive that “Section 6751(b)(1) includes no requirement that all potential penalties be initially determined by the same individual nor at the same time.” 152 T.C. at 85. Professor Keith Fogg notes in his Procedurally Taxing Blog write-up of the Roth case that this 10th Cir. opinion solidifies the case law that permits Chief Counsel attorneys to assert new penalties. What he could have added was that it does so by reading §6662 as imposing more than one penalty. Note, however, it did not say Chief Counsel can cure a defectively proposed penalty by reasserting the penalty in Tax Court. Chief Counsel might in fact be able to do so, but that’s another issue for another time.
Third, when you look at the actual error committed by the taxpayers for which the IRS sought the 40% penalty you see how lucky they were to find a procedural dodge to the penalty bullet.
The Schedule A error was an improper charitable contribution. The taxpayers were lured into participating in a charitable donation scheme promoted to them by their long-time accountant and return preparer. No, it was not a conservation easement. The scheme was something almost straight out of The Honeymooners. It promised that if they (a) bought 3,400 designer eyeglass frames sight unseen for $50,000 and then (b) donated the frames one year later to a charity who would use them as part of its charitable purpose, they would be able to take a $225,000 charitable donation. Assuming they were in the 39% marginal tax bracket, and assuming my math skills are adequate, that would reduce taxes by something like $88,000.
The unbaked part of this scheme was that the 3,400 frames they were supposedly buying were sitting in some warehouse mingled with some 166,600 other frames that 49 other suckers taxpayers were also buying in similar lots of 3,400 at $50,000 per lot. 50 lots. The supposedly qualified appraisal they obtained, however, was of this whole kaboodle of 170,000 frames. It pegged the values of the frames from between $37 to $80 per frame, and said it all added up to about $11-12 million. Divide by 50 and you get about $225,000. A later appraisal said 40,000 of the frames were in fact worthless. Neither of the appraisals attempted to identify or value just the 3,400 frames the Campbells had purchased for the purpose of making a donation. That proved fatal to their attempted deduction.
The cherry on top of this half-baked pie was this: when the Campbells eventually received the donation receipt from the charity, the receipt failed to comply with the §170(f)(8)(B) requirement that it tell the donor what goods or services had been given in exchange for the donation. It had no language that said what goods or services the taxpayer had received in exchange. It said nothing. Was. Just. Silent.
Wow. Who would file a return with such a patently defective receipt? And the Campbells even had a return preparer who you would think would have known better. Once the matter got to Tax Court the Campbell’s attorneys tried an ex-post Hail-Mary substantial compliance argument. It went about as well as you would expect. But the Campbell's attorneys did well to catch the defective penalty approval form and grab a partial win for their clients. They knew the lesson that we all now know: §6662 contains two penalties, not one, for penalty approval purposes.
Bryan Camp is still the George H. Mahon Professor of Law at Texas Tech University School of Law, even though he now has to work from his home.