We welcome back guest blogger Bryan Camp who is the George H. Mahon Professor of Law at Texas Tech. Professor Camp teaches both tax and administrative law which is why I sought him out for this guest post. The decision here is important. The lead attorney for the taxpayer, Jerry Kafka, is one of the best if not the best tax litigators in the country. Though his client lost this case the arguments made here were not frivolous. What could have been a game changer had the taxpayer won leaves us in the same posture we were in before the case was brought but with more light shone into the corners of tax and the APA. Keith
Keith emailed me last week and asked if I would care to blog about a recent 4th Cir. opinion affirming a Tax Court decision that upheld a proposed deficiency in the taxes of QinetiQ US Holdings (Q). (for previous PT posts on QinetiQ see here, here and here).It seems that Q took a big §83(h) deduction. On audit, the Service disallowed the deduction and sent Q a Notice of Deficiency (NOD). In court, Q argued that the NOD violated the Administrative Procedure Act (APA) because the NOD gave no explanation for the disallowance and, oh, by the way, the §83 deduction was proper. The Tax Court rejected both arguments. The 4th Cir. affirmed.
Maybe the §83 issue is interesting. If so, I’m sure the Surly Subgroup will blog it. To me, however, what makes this 4th Cir. opinion worthy of a shout out is its discussion about the relationship of the Administrative Procedure Act (APA) to tax procedure. Ever since the Supremes decided Mayo Foundation in 2011, it seems everybody and their little dogs have been declaring that something called “tax exceptionalism” is dead. The Fourth Circuit’s opinion gives a more nuanced take, one that is worth blogging about for three reasons. First, it represents a new front on the “tax exceptionalism” debate. Second, the Circuit’s opinion makes a critically important point about the relationship of the APA to tax procedure. Third the opinion could affect court review of other types of IRS determinations, such as CDP determinations.
I will consider each of these points in turn.
- A New Front on Tax Exceptionalism Debate
QinectiQ represents a new front in the ongoing debate over the proper relationship of the APA to tax practice and procedure. Up to now, the debate has been chiefly about tax regulations and tax guidance. That’s the Mayo case and other cases where taxpayers have sought to challenge the procedure by which Treasury and the IRS have issued regulatory and similar guidance. This case involves the proper application of the APA to a very different type of agency action: an agency determination. The APA has some general standards that courts are supposed to use in reviewing agency determinations in particular cases, also known as “agency adjudications.”
Professor Steve Johnson at Florida State has written extensively and lucidly about the tax exceptionalism debate. In this short Florida Bar Review article from 2014 he encouraged tax practitioners to consider challenging NOD’s under an APA standard. Apparently the lawyers for Q read his article!
The APA is found at 5 U.S.C. Subchapter II. Section 706 says that courts should review agency adjudications to be sure they were not made arbitrarily. To do that, the court needs to see what the agency’s rationale was for its decision. So over time the Supreme Court has developed the requirement that “an agency provide reasoned explanation for its action.” FCC v. Fox Television, 556 U.S. 502, 515 (2009).
In QinectiQ, the taxpayer argued that the NOD failed the APA §706 standard. The NOD said only that Q had additional taxable income of “$177,777,501” because Q had “not established that [it was] entitled” to a deduction “under the provisions of [26 U.S.C. §83].” The NOD gave zero explanation for why the Service was disallowing the deduction. The failure to articulate a rational explanation for its disallowance decision meant that a review court had no way to police the NOD for arbitrariness.
In a short unpublished order, the Tax Court refused to apply the APA standard and held that the NOD was instead subject to the standard provided for in §7522(a). The taxpayer’s argument to the Fourth Circuit was essentially that §7522 and the APA standards were cumulative, not exclusive.
The Fourth Circuit affirmed the Tax Court. It believed the taxpayer’s attempt to apply the APA standard “fails to consider the unique system of judicial review provided by the Internal Revenue Code for adjudication of the merits of a Notice of Deficiency.” (p. 9 of slip opinion).
The Fourth Circuit thought two features of tax administration made the APA standard inapplicable. First, “because the Code’s provisions for de novo review in the tax court permit consideration of new evidence and new issues not presented at the agency level, those provisions are incompatible with the limited judicial review of final agency actions allowed under the APA.” (p. 10-11 of slip opinion). Second, the Tax Code’s provisions for judicial review of NOD’s pre-dated the APA. “Congress did not intend for the APA ‘to duplicate the previously established special statutory procedures relating to specific agencies.’” (p. 12 of slip opinion, quoting Bowen v. Massachusetts, 487 U.S. 879, 903 (1988)).
The Fourth Circuit’s consideration of these two features of tax administration is the more nuanced understanding that I think is worth commenting on.
- The Proper Relationship of the APA to Tax Administration
The nuance is this: the APA is not sui generis. That is, the APA was enacted on top of existing agency practices and procedures. One simply cannot pretend that the APA was enacted in a vacuum! That’s the point I try to make about tax regulations in my article “A History of Tax Regulation Prior the Administrative Procedure Act,” 63 Duke L. J. 1673 (2014)
The APA was enacted on the basis of a massive, massive, study of federal agencies and their operations undertaken by the Attorney General’s Committee on Administrative Law (“the Committee”). The Committee’s Final Report is generally believed to be the most important influence on the text and application of the APA.
The Final Report grew out of a detailed study of then-existing agencies, a study contained in 27 Monographs written by staff, each running hundreds of pages. (Monograph 22 focused on the tax administration). At its inception, the Committee “had initially hoped to be able to suggest uniform rules for agency practice” (quote from Grisinger Law in Action: The Attorney General’s Committee on Administrative Procedure, 20 J. of Policy History 379 (2008)). In light of the information produced in the 27 monographs, however, the Final Report backed away considerably from that aspiration and instead prescribed a general framework for balancing the goals of agency efficiency and autonomy with the goals of agency transparency and protection of individuals from arbitrary agency actions. That is why the resulting APA was widely understood as standing for the proposition that “procedural uniformity was not well suited to the administrative process.” (Grisinger at 402; one sees the same theme in almost all the contemporary commentaries and reviews of the Final Report). That is, the APA provided generalized standards for controlling administrative actions rather than detailed prescriptions. This conventional view is elegantly summed up by Professors Hickman and Pierce: “the Administrative Procedure Act is to administrative law what the Constitution is to constitutional law.” Kristin E. Hickman, Richard J. Pierce, Jr., Federal Administrative Law: Cases and Materials, (Foundation Press, 2010) at 19.
What this means is that while the APA does apply to all agencies, including the IRS, it does not apply in the exact same way to all agencies. Every agency is “exceptional” in that every agency faces a different set of operational demands and requirements and organic statutory provisions. All of those variables must be reconciled to the general language of the APA and it should not surprise anyone that different reconciliations lead to different applications of the APA principles to different agencies. That is why the Supreme Court, in Bowen, said “When Congress enacted the APA to provide a general authorization for review of agency action in the district courts, it did not intend that general grant of jurisdiction to duplicate the previously established special statutory procedures relating to specific agencies.” 847 U.S. 879 at 903.
Put another way, the debate is not “whether” the APA applies, it’s “how” the APA applies. It is not so much whether the NOD review procedure “comply with” the APA as it is whether the procedures are “consistent with” the APA. Does the APA displace or otherwise affect otherwise applicable rules that govern what goes into the NOD and how the Tax Court reviews it?
That is what the Fourth Circuit recognizes in QinetiQ. The Tax Code’s specific statutory review structure makes the APA review standard inapplicable, for both historical and operational reasons. The historical reason is what I said above: the specific statutory structure for court review of NOD’s pre-dates the APA and the APA was not written to displace prior law. The operational reason is that taxpayers have the burden to prove entitlement for deductions and have every opportunity to do so in a de novo Tax Court review. That de novo nature of review is what makes the current practice acceptable. For example, if the IRS rejects a claimed deduction, tax law does not put the burden on the IRS to prove up the rejection. The burden remains on the taxpayer to prove up the entitlement, only now in front of the Tax Court (or district court if the taxpayer chooses to pay the deficiency and then go for a refund). It is the Tax Court’s job to determine or re-determine the taxpayer’s proper tax liability. That’s why it can either increase the proposed deficiency (§6214(a)) or actually order a refund (§6512(b)).
The Tax Court has recognized these points as well. It has a nice discussion of this kind of “tax exceptionalism” in Ax v. CIR, 146 T.C. No. 10 (2016) (which Les has previously blogged here and which Professors Stephanie Hoffer and Chris Walker give some very thoughtful comments here). In Ax, the taxpayer objected to the Service raising a new issue before the Tax Court, even though the Service acknowledged it bore the burden of proof. Like the taxpayer in QinetiQ, the taxpayer in Ax argued that because “the Supreme Court rejected the concept of ‘tax exceptionalism,’ the Administrative Procedure Act and [case law] bar Respondent from raising new grounds to support his final agency action beyond those grounds originally stated in the notice of final agency action” (e.g. the NOD). The Tax Court’s rejection of that position is worth reading.
III. The Door Is Still Open: Implications of QinetiQ on Other IRS “NODs”
Have you ever noticed how you need an NOD to get Tax Court review? I don’t just mean the “Notice of Deficiency.” I also mean the “Notice of Determination” from a CDP hearing. That’s a ticket to the Tax Court, too. But, sorry, a “Determination Letter” is not a ticket. Likewise, if a taxpayer petitions for “stand alone” spousal relief per §6105(f), the eventual “Notice of Determination” issued by IRS or Appeals is the ticket for Tax Court review (of course, §6105(e) also permits a taxpayer to seek judicial review in cases where the Service has not acted within 6 months of the initial request for spousal relief).
The point is that the Tax Court reviews agency decisions other than deficiency determinations. QinectiQ deals with only ONE kind of IRS determination (although by far the most frequent). The inimitable Steve Johnson gives an excellent and in-depth treatment of the variety of ways that the APA §706 might be applicable to a variety of IRS determinations in his Duke L. Rev. article “Reasoned Explanation and IRS Adjudication,” 63 Duke L. J. 1771.
The Fourth Circuit’s rationale for not applying the ABA §706 standard of review in QinectiQ actually suggests the ABA standard may be applicable to court review of some of these other IRS determinations. One sees this in the opinion’s discussion of Fisher v. Commissioner, 45 F.3d 396 (10th Cir. 1995). In Fisher, the 10th Circuit held an NOD invalid because the NOD implicitly denied, without explanation, a taxpayer’s request for penalty abatement. Since the Service has the discretion to grant or deny such requests, the 10th Circuit thought that the failure to explain why the Service was exercising its discretion to deny the relief violated “an elementary principal of administrative law that an administrative agency must provide reasons for its decisions.” 45 F.3d at 397. Unexplained exercise of discretion is per se arbitrary, says Fisher.
The Fourth Circuit could have just disagreed with Fisher. The IRS issued a well-reasoned AOD that explained why Fisher was wrong. AOD-1996-08, 1996 WL 390087. But the Fourth Circuit instead chose to distinguish Fisher, saying in footnote 6: “we do not read Fisher…as requiring a reasoned explanation in all Notices of Deficiency.” Hmmmm. Does that suggest that in situations where the Service is exercising discretion—like refusing to grant a request for spousal relief, or refusing to accept a collection alternative offered in a CDP hearing—that one of those decisions would be subject to the APA §706 standard, even when the Tax Code has very detailed special statutory procedures? After all, both the CDP provisions and spousal relief provisions were added by Congress after the APA.
Let’s look at CDP procedures. Currently the Tax Court’s approach to CDP review is both (a) abuse of discretion and (b) de novo. That’s not quite square with how the APA contemplates the relationship of a reviewing court to agency decisions. Here’s how the Court explained it in a recent CDP case, Drilling v. Commissioner, T.C. Memo 2016-103:
the standard of review employed by the Tax Court is abuse of discretion, except with respect to the existence or amount of the underlying tax liability, for which the standard of review is de novo. Goza v. Commissioner, 114 T.C. 176, 181-182 (2000). The evidentiary scope of review employed by the Tax Court is de novo. Robinette v. Commissioner, 123 T.C. 85, 101 (2004), rev’d, 439 F.3d 455, 459-462 (8th Cir. 2006). That means that the Court’s review is not confined to evidence in the administrative record. See Speltz v. Commissioner, 124 T.C. 165, 177 (2005) (citing Robinette v. Commissioner, 123 T.C. at 94-104), aff’d, 454 F.3d 782 (8th Cir. 2006). If the Court remands a case to the Appeals Office, the further hearing is a supplement to the original hearing, not a new hearing, Kelby v. Commissioner, 130 T.C. 79, 86 (2008), but the position of the Appeals Office that the Court reviews is the position taken in the supplemental determination, id.
Notice that this means if the taxpayer wants to present new information, the Tax Court has the option of hearing that new evidence itself or sending the case to the Appeals Office for a “supplemental” hearing. See e.g. Drake v. Commissioner, T.C. Memo 2006–151, aff’d 511 F.3d 65 (1st Cir. 2007) (“The resulting section 6330 hearing on remand provides the parties with the opportunity to complete the initial section 6330 hearing while preserving the taxpayer’s right to receive judicial review of the ultimate administrative determination.”)
The Tax Court’s practice of allowing new information is IMHO a perfectly reasonable procedure and it reflects the ongoing nature of both CDP and 6015(f) determinations. Each of those determinations can be affected by facts that change at any time. But it is arguably NOT the procedure contemplated by the APA. Notably, the APA contemplates that the record, once made, is unalterable. And the danger of allowing an open record is that the Tax Court becomes mired “with tax enforcement details that Congress intended to leave with the IRS.” Robinette v. Commissioner, 439 F.3d 455, 459 (8th Cir. 2006) aff’ing in part 123 TC 85.
Both the CDP and the spousal relief review provisions were added by Congress long after the APA’s enactment. Perhaps the flip side of pre-existing administrative schemes not being displaced by the APA is that post-APA statutory provisions do not exclude application of APA §706 but incorporate that standard (unless of course Congress says the provisions are to be exclusive). Of course, the operational reasons for concluding that the §706 standard has been trumped by the specific CDP provisions may remain.
Those of us who study this stuff are not in agreement. For Les’ take, see here; for Stephanie Hoffer and Chris Walker’s take, see here. As Keith points out, the CDP procedures have astonishingly large gaps in them. But IMHO the APA does not mandate a uniform set of rules for the Tax Court to deal with those gaps. Like the U.S. Constitution, the APA simply provides the touchstone by which to measure any rules or procedures that the Tax Court or IRS come up with in implementing CDP. Claiming that a procedure violates §706 is like claiming one process or another violates “due process.” You first have to figure out what process is “due” before you can find a violation.
In sum, I believe the Fourth Circuit’s opinion in QinetiQ leaves open the door to argue that for non-deficiency determinations, APA §706 has greater applicability than for standard Tax Court review of deficiency notices. Personally, I think that (1) the specificity of the both the CDP and innocent spouse provisions, and (2) the specific relationship that the Tax Court has in supervising so many aspects of tax administration still trump the general provisions in the APA. But those two reasons for treating current procedure as may not be as applicable to other types of determinations, such as §6672 decisions, or penalty abatement decisions, or other “discretionary” decisions that are not clearly covered by specific Tax Code provisions.