Paul L. Caron

Monday, January 22, 2018

Lesson From The Tax Court: Treasury Regulations And The APA

No doubt there is a lot of dirty bathwater in the Treasury Regulations, codified in title 26 of the Code of Federal Regulations (CFR). The upside of the current administration’s anti-regulation focus is that it is allows Treasury to prioritize scrubbing unneeded regulations. Treasury reported on its progress in October noting that “the IRS Office of Chief Counsel has already identified over 200 regulations for potential revocation, most of which have been outstanding for many years.”

To be sure, it’s a small upside. Some regulations become outdated because they are simply overtaken by statutory changes. For example, Treas. Reg. 1.217-2(b)(1) allows taxpayers to deduct the cost of meals when moving to start a new job. That was fine under the statute Congress originally enacted in 1969, but it became obsolete when Congress modified the statute in 1986 to specifically disallow meal expenses as a deductible item. And now, of course, Congress has repealed the moving expense deduction entirely, but the regulations will still be there.

Other regulations become outdated because of societal change. My favorite example is former Treas. Reg. 1.162-6 which started off this way: “A professional man may claim as deductions the cost of supplies used by him....” To modern eyes, that regulation obviously denied deductions to taxpayers not in the trade or business of being a “professional man” ...such as anyone who was only a man as a hobby and not as profession. Think Victor, Victoria. Treasury nuked that reg in 2011.

The scrubbing effort carries a small upside because outdated regulations generally do little harm. I tell my students that is why you have to read the actual statutory language first. In real life, of course, tax practitioners rely on the commercial services like BNA, CCH or RIA to summarize the rules and those services keep current. Taxpayers reporting their 2017 taxes are unlikely be blindsided by the moving regulations into trying to deduct meal expenses in a move. Likewise, taxpayers reporting their 2018 taxes are unlikely to try and deduct moving expenses at all, much less in reliance on the regulations.

But the focus on throwing out the bathwater presents an obvious danger to the baby. The ham-fisted 2-for-1 requirement of Executive Order 13711 is not just focused, it’s myopic. Another danger is posed by the myopic thinking that the word “regulation” has the same meaning for all agencies and that the Administrative Procedure Act (APA) applies in lock-step to all agencies. Both myopias ignore the vast difference in purpose of regulations issued by different agencies.

Last week’s Tax Court opinion in SIH Partners LLLP, et al. v. Commissioner, 150 T.C. No. 3, January nicely illustrates the purpose and use of tax regulations. In it, the taxpayer tried to invalidate a 45 year old regulation for failing to meet APA requirements. The Tax Court has a nice opinion applying the APA with sensitivity to the tax regulation process and suggests a clearer view of what makes tax regulations different from those of many other agencies.

More below the fold.

The case of SIH Partners LLLP basically involves five guys---Eric Brooks, Joel Greenberg, Arthur Dantchick, Jeffrey Yass, and Mark Dooley---who run an investment business through a company called Susquehanna Investment Group, LLC (SIG). Each of the five guys was also a 100% shareholder of a separate S corp. Together these five S corps owned 99% of another entity: Susquehanna International Holdings Partners LLLP (“SIHP”). In tax years 2007 and 2008 the Tax Court found that SIHP was itself a 100% shareholder of three foreign corporations: SIHL, SEHL and STS. The “S” in all these foreign corporations stands for Susquehanna.” Natch. So SIHL was Susquehanna Ireland Holdings Limited (Ireland), SEHL was Susquehanna Europe Holdings Limited (Luxembourg) and STS was Susquehanna Trading Services, Inc. (Cayman Islands). The Tax Court found that all three of these entities were Controlled Foreign Corporations of SIHP during 2007 and 2008.

In 2007 SIG borrowed over $1.485 billion from Merrill Lynch. Of that amount $1.285 billion remained outstanding as of December 231, 2008. The reason those dates matter is because during that time all three CFCs signed on as guarantors of all of the notes securing all those loans. And, during those years, each of the CFCs had a bunch of earnings and profits but did not distribute them to SIHP. For example, as of January 1, 2007, STS has accumulated earnings and profits of over $84 million. As of January 1, 2008, that figure went up by the $1.4 million STS earned in 2007. STS earned another $1.6 million more in 2008 but, again, did not distribute anything to SIHP.

The issue in the case was whether SHIP had to include in income those earnings of its three CFCs because of guarantees those CFCs had made on the Merrill Lynch loans to SIG. Section 951, by cross-reference to §956, requires a taxpayer like SIHP to include in gross income the lesser of (1) its share of the average amount of “United States property” held by their CFCs, or (2) its share of the CFC’s “applicable earnings.” Section 956 gives a bunch of rules for what constitutes “United States Property.” One rule is that “United States Property” includes the obligation of a United States person. As part of that rule §956(d) says that a CFC “shall, under regulations prescribed by the Secretary, be considered holding an obligation of a United States person if...[the CFC] is a pledgor or guarantor of such obligation.”

Congress enacted §956 in 1962. In April 1963 Treasury issued proposed Regulations interpreting the scope of the statutory command in §956. It received comments and issued final regulations in February 1964.

You see where this is going? SIG was a United States person. It was obliged to Merrill Lynch for over $1 billion. The CFCs were guarantors of the obligation. Applying Treas. Regs. 1.956-1 and 1.956-2 in this case, the IRS determined that SIHP had to report all of the CFC’s applicable earnings.

End of story, right? Nope. SIHP argued to the Tax Court that (1) the regulations were invalid because Treasury failed to comply with the Administrative Procedure Act (APA) and (2) absent valid regulations §956(d) had no force or effect of law. I will comment on each argument in turn.

The first argument was that the regulations were invalid. The Tax Court ruled otherwise. First the Tax Court agreed with SIHP that the regulations were “legislative” regulations and thus the APA notice and comment requirements applied. And the Treasury had used the notice and comment procedure. But the APA requires agencies to do more than just give notice. 5 USC §553(a) requires that “After consideration of the relevant matter presented, the agency shall incorporate in the rules adopted a general statement of their basis and purpose.” (emphasis supplied). The Supreme Court has interpreted that statutory requirement to mean that agencies must give an adequate basis and explanation for the regulation. Motor Vehicle Mfrs. Ass’n of the U.S. v. State Farm Mutual Auto Ins. Co., 463 U.S. 29, (1983) (“State Farm”).

Here’s the explanation given in Treasury Decision 6704, 29 Fed. Reg. 2591, 2599 (Feb. 20, 1964): “After consideration of all such relevant matter as was presented by interested persons regarding the rules proposed, the amendment containing the regulations under section 956 is hereby adopted.”

SIHP argued that this explanation was---to use the current administration’s favorite term---a “nothingburger.” SIHP said this vague language failed to satisfy the Supreme Court’s interpretation of an agency’s duties in the State Farm case.

The Tax Court disagreed and does a really nice job in explaining that the Court’s task is to evaluate the entire process of a rule’s promulgation to evaluate compliance with the APA. What was important for the Tax Court here was that Treasury had nothing to comment on. It had received no substantive comments during the notice and comment period. The Court concludes:

“The administrative record reflects that no substantive alternatives to the final rules were presented for Treasury’s considering during the rulemaking process. The agency did not act arbitrarily or capriciously by failing to address contrary viewpoints or finding of fact that were never developed or presented.”

I see at least three lessons in this case.

First, the case shows that Tax Court takes it job seriously to review Treasury’s compliance with the APA. One sees this as well in last week’s Kasper v. Commissioner, 150 T.C. No. 2. Les Book has a great post on Procedurally Taxing on Judge Holmes’ very thorough opinion on how the Tax Court applies its APA review powers in a variety of contexts. As with its evaluation of the Treasury regulation in this week’s case, the Tax Court in Kasper shows how it applies the APA with sensitivity to the context of tax administration.

The second lesson is about that sensitivity. Nah, I don’t mean feelings. We’re tax lawyers! I mean awareness, understanding, and willingness to consider that administering the tax laws is quite a bit different than administering other laws--- safety standards for cars, for example.  Maybe I should put it this way: a mental flexibility to be open to the idea that tax regulations are different from regulations issued by other agencies.  Now I do not mean that neither Treasury or the IRS is subject to the APA.  No one has ever made that claim to my knowledge. But to think that the APA is a straight-jacket is misguided formalism. As a friend of mine says: “When you have a hammer, everything looks like a nail.” Those who think the APA is a hammer tend to want to say all agency actions are alike. A regulation is a regulation is a regulation. Same same.

I think it more useful to think of the APA as a Swiss Army knife. This case shows why.

The taxpayer here wanted to say that the 45 year old Treasury Regulations were invalid because the agency had not given a reasonable enough explanation. The taxpayer brought up all sorts of interesting arguments about the subtleties of guaranties and I think was pretty convincing that the §956 regulations did not address what it termed “economic realities.” The taxpayer pointed to some pretty powerful language in State Farm supporting its contention that Treasury’s nothingburger preamble failed the APA’s reasoned explanation requirement.

The Tax Court rejected using the APA as a hammer. It found that the regulations at issue in State Farm differed materially from the §956 regulations in two respects. First, the procedural posture of the regulatory action was quite different. State Farm concerned regulations issued by the National Highway Transportation Safety Administration (NHTSA). The regulations at issue there had rescinded prior regulations that required car makers to install passive restraints in cars (either automatic seatbelts or airbags). Those prior regulations, in turn, had been the product of more than 10 years of development. As the Supreme Court noted: “Over the course of approximately 60 rulemaking notices, the requirement has been imposed, amended, rescinded, re-imposed, and now rescinded again.”

That difference in procedural context made a difference to the Tax Court:

“the facts surrounding the agency action in State Farm are inapposite, and the concerns that led the Court in that case to review more stringently the agency’s process are not presented here...The regulations at issue dd not reverse previously settled agency policy, and they were not promulgated contrary to facts or analysis that supported a different outcome."

Second, the Tax Court recognized a critical substantive difference between the passive restraint regulations at issue in State Farm and the §956 regulations at issue in SIHP.

"Treasury’s rulemaking in this case differs fundamentally from the agency action in State Farm because Treasury’s decision did not (can could not) purport to rely on findings of fact. Treasury’s actions promulgating the rules for pledges and guaranties reflect at most the implementation of a policy judgment..."

In other words, dear readers, Treasury was doing what Courts do: interpreting law. Court interpret law by looking at text, context, and purpose. Court decisions about the meaning of tax statutes reflect “at most...a policy judgment” on how best to effectuate the law Congress wrote.

There is another, perhaps more subtle but equally important difference between the State Farm regulations and the §956 regulations. The car regulations at issue in State Farm were what we call “command and control” regulations. They imposed constraints on how car makers could do their business. Car makers had to install passive restraints. Car makers would violate the law if they did not obey the regulations. In contrast, the §956 regulations did not impose any requirements on how taxpayers here conducted their investment business. They were free to have their CFCs make pledges or not, as their business needs dictated. The regulations simply told the taxpayers what the tax consequences of certain structures would be. The taxpayers broke no laws by choosing their preferred financial structure.

I said this was a subtle difference because I can hear some readers grumble that tax regulations influence behavior just like command-and-control regulations do. I agree that if one is looking simply at how regulations affect behavior, then perhaps some tax regulations affect behavior as much as command-and-control. But even when they do, they still always do so indirectly, by setting up tax consequences.  It is up to the taxpayer to decide how those consequences balance with other factors.  Car makers had no such choice.  That is, I submit, an important difference that should affect how we evaluate tax regulations and their creation under the APA.

The final lesson is what issue the Court stays silent on. The Court did not address the argument raised by the taxpayer SIHP that if these regulations were indeed invalid, then the income inclusions determined by the IRS could not sustained. The idea behind SIHP’s argument is that the language Congress used in §956 makes the section inoperative unless and until Treasury enacts regulations. The Tax Court did not need to address this idea in light of its holding that the contested regulation was valid. But the idea that a revenue statute cannot be applied by the IRS absent Treasury regulations is an idea worth some further exploration.

Here is again the statutory language: §956(d) says that a CFC “shall, under regulations prescribed by the Secretary, be considered holding an obligation of a United States person if...[the CFC] is a pledgor or guarantor of such obligation.” The idea is that the statute reads a CFC “shall be considered holding an obligation ONLY if the Secretary prescribes regulations.”

The idea that §956(d) is not, in some sense “self-executing” is difficult to square with three practicalities of tax administration. First, all taxing statutes are self-executing...and at the same time none are! That is, they are all implemented by taxpayers who self-report both their prior transactions and the taxes they owe. The idea that a tax statute is not “self-executing” ignores this unique feature of tax law.  At the same time, while our system of taxation relies upon taxpayers to report and pay tax upon already completed transactions in the first place, the IRS is always there to Monday-morning quarterback the tax positions taken by the taxpayer. Car makers cannot make cars in compliance with the statute until the agency tells them what constitutes an adequate safety feature.  But taxpayers can always report their taxes.  If the IRS has not told them upfront how to report income from CFCs who have guaranteed a loan, the IRS can always tell them on audit.  

The second practicality is that reporting and auditing are both exercises is looking back and fixing errors that happened in the past. That inherently retroactive feature of tax administration is explored in a thoughtful paper by Professor James Puckett, “Structural Tax Exceptionalism,” 49 Ga. L. Rev. 1067 (2015). Whether a CFC’s guarantee is enough to be holding an obligation of a U.S. taxpayer is simply a question of interpreting the meaning of the statutory language as applied to completed transactions and that interpretation then informs taxpayer decisions on how to structure future financial transactions. To read a standard delegation of authority to flesh out a statute’s meaning as a prohibition on applying the statute until a regulation come out interpreting the statute, runs counter to this retroactive aspect.

Finally, the idea that neither taxpayers nor the IRS can apply the statute without an implementing regulation runs contrary to a fundamental principal of administrative law: The choice on whether to use regulations or use case-by-case adjudications to implement a statute “lies in the first instance within the [agency's] discretion." NLRB v. Bell Aerospace Co., 416 U.S. 267, 294 (1974). That is, the IRS should be able to choose to apply a given statute by simply auditing returns and taking a position or by writing a regulation.

With over 155 million individual taxpayers, it is not surprising that the IRS chooses to try and give broad guidance by regulation as much as possible. Tax regulations do not create rights or burdens, etc., so much as they explain to taxpayers how taxpayer are to discharge the duties Congress has created in the statutory language. The myopia inherent in either the notion that tax regulations are “bad” or else the equally pernicious notion that tax regulations “legislate” rather than “interpret” statutes both may lead to unintended, and not salutary, consequences, some of which Professor Puckett suggests.

It will be interesting to see how this last idea of “we get away with ignoring the statute because you did not write a regulation” fares in future cases.   I hope it gets soundly rejected.  Taxpayers need tax guidance and tax regulations give the guidance and certainty that taxpayer can rely on.   Even when those regulations are mistaken in some sense, it is usually better to have a bad bright line rule than no rule at all or a rule that varies from Revenue Agent to Revenue Agent.

Bryan Camp, IRS News, New Cases, Tax, Tax Practice And Procedure | Permalink


It sounds like this is another good Tax Court decision. An interesting post might be a list of top five Worst Tax Court decisions. I don't mean being reversed on appeal, but on what *ought* to have been the holding, since appellate courts are probably worse on tax.

Posted by: Eric Rasmusen | Jan 22, 2018 8:30:56 AM

This raises a deference question. If the agency enacts a regulation with notice and comment but no commenter addresses problem X so the agency doesn't think about it, should the agency still get deference from the court on point X? I suppose that's been addressed already; it sounds like it would be at the heart of Chevron deference.

Posted by: Eric Rasmusen | Jan 22, 2018 8:29:07 AM