We welcome back guest bloggers Susan C. Morse and Stephen E. Shay. They bring us a further update on the efforts of the taxpayer in the Altera case to have the Supreme Court accept the case for argument. Keith
Previously we blogged here (crossposted at Yale JREG Notice & Comment) about the government’s May 14 brief in opposition to the taxpayer’s petition for certiorari in Altera v. Commissioner. On June 1, Altera replied to the government’s brief, as explained here by Chris Walker. The case has been distributed for a Supreme Court conference later in June.
The Altera reply brief doubles down on an argument that the government brief has already persuasively dispatched: that Treasury gave the impression during the rulemaking process that comparability analysis – i.e., the analysis of comparable transactions between unrelated parties – was relevant to the determination of an arm’s length result under the transfer pricing regulation at issue, and that then the government changed its tune.
read more...First, some background to level-set for any new readers. In its cert petition, the taxpayer asked the Supreme Court to review a Ninth Circuit decision upholding a 2003 amendment to an existing tax regulation governing intra-group cost-sharing arrangements for the development of intangible property. (We submitted amicus briefs on behalf of the government to the Ninth Circuit in earlier stages of this litigation here (with coauthors Leandra Lederman and Clint Wallace), here and here.)
The regulation conditions the benefits of a qualified cost sharing arrangement, or QCSA, on including stock-based compensation deductions related to developing intangible property in the pool of costs to be shared. If this (and other) QCSA conditions are met, the cost-sharing party — typically an offshore subsidiary of a U.S. multinational firm — owns a share of the rights in intangible property, even though this intangible property is often developed within the United States. Allowing an offshore subsidiary to own a share of intangible property means that a U.S. multinational firm can attribute some profit from intangibles to the offshore subsidiary. This in turn means that the U.S. multinational firm can avoid paying U.S. corporate income tax on some of its profit.
Altera proposes that the Supreme Court should take this case because it is an opportunity to place limits on an inappropriate exercise of administrative agency power. The taxpayer’s cert petition argues that Treasury did not provide a reasoned explanation for the regulation as required under State Farm, in light of evidence cited by commenters that unrelated parties to similar types of arrangements did not share stock-based compensation costs; that the government in litigation engaged in post hoc rationalization to defend the regulation, in violation of Chenery I; and that the Ninth Circuit accorded Chevron deference to a procedurally defective regulation.
The government in response observed that the taxpayer conflates the arm’s length standard with comparability analysis. It explained that the government has maintained a consistent argument throughout the rulemaking process and this litigation. That is, the government has consistently maintained that the 2003 regulation’s rejection of comparability analysis as a means of determining an arm’s-length result in this limited context is consistent with both the “commensurate with income” language of the statute adopted in 1986 and the accompanying legislative history.
The core of Altera’s argument is that the government surprised taxpayers and tax advisers by making a “sea change in tax law without providing any notice of the change or opportunity to comment on it” (Reply Br. 1) and by taking a “new position” in litigation (Reply Br. 2) about the meaning of the arm’s length standard. Altera’s reply brief states this claim in at least three ways. None hold up.
The first thing Altera claims is that “The arm’s length standard has a settled meaning: A transaction meets the arm’s length standard if it is consistent with evidence of how unrelated parties behave in comparable arm’s length transactions.” (Reply Br. 5) Altera may wish that this sentence stated doctrinal transfer pricing tax law, but it does not. As the government’s brief in opposition to the cert petition correctly explains, Altera’s statement conflates the arm’s length standard with comparability analysis. Comparability analysis is not a predicate for determining an arm’s length result. One clear indication of that reality is the residual profit split transfer pricing method contained in regulations promulgated in 1994.
The second claim Altera makes is that the government initially suggested that comparability analysis is relevant to the determination of an arm’s-length result under the regulation at issue in this case, but then changed its mind. This is also incorrect. As the government’s brief explains, Treasury promulgated the 2003 amendment to make explicit what it had consistently argued was implicit in the prior (1995) cost-sharing regulation: that QCSA stock-based compensation costs must be shared to produce an arm’s-length result, without regard to evidence of allegedly comparable transactions. And it consistently pointed to the commensurate-with-income language of the statute and the related legislative history to support its position. It referred to commensurate-with-income both in the 2002 Notice of Proposed Rulemaking and in the 2003 Preamble.
This government’s position in this regard has been at the heart of a longstanding and well-known disagreement between taxpayers and the government. In 2002, lawyers at Baker & McKenzie explained the already-long history, in a comment to the proposed regulations written on behalf of Software Finance and Tax Executives Council:
On audit, in Advance Pricing Agreement negotiations, in docketed Tax Court cases, in published field service advice, and in speeches by Service officials … the Service has taken the position that stock-based compensation … must be included in related parties’ cost sharing pools. … Taxpayers have steadfastly and vehemently disagreed[, … absent] any evidence that unrelated parties … share stock option “costs” in their own cost sharing pools.
The third claim that Altera makes is that taxpayers did not realize that the government was promulgating a rule that did not rest on comparables and were caught by surprise. It writes that “none of the companies, industry groups, or tax professionals that participated in the rulemaking noticed” (Reply Br. 2) that the 2003 amendment made evidence of allegedly comparable uncontrolled transactions not determinative of an arm’s length result in this context. This claim also does not hold up. Indeed, the amended regulation itself – in both its proposed and final form – unequivocally states that a QCSA will achieve an arm’s-length result “if, and only if,” the parties share all development-related costs (including stock-based compensation costs) in proportion to anticipated benefits.
In written submissions and at the 2002 hearing to consider the proposed regulation, commenters certainly realized that the regulation was not based on evidence of comparables. A representative for the American Electronics Association stated that the regulation identified an arm’s-length result “by fiat,” implicitly acknowledging that the government had rejected a comparables-based inquiry. A Fenwick & West partner explained that the regulation “deem[ed] a result to be arm’s length without providing any evidence.” A tax partner at PricewaterhouseCoopers noted the perception that the amendment “seem[s] contrary to the arm’s length standard as evidenced by actual transactions ….” The rest of the regulatory record is consistent. Commenters understood. Taxpayers and tax advisers knew exactly what Treasury was doing.
Altera says it is making an administrative law argument, but it is really interested in a tax policy outcome. The asserted “immense prospective importance” (Reply Br. 4) is illusory. Even if the Court were to grant the petition and then hold that the 2003 amendment is procedurally defective, Treasury could simply re-promulgate the rule without substantive change but with a more detailed explanation. As for past tax years, Altera’s and similarly-situated companies’ financial statements have already incorporated the possibility that corporate income tax will be due based on compliance with the regulation. The real importance of the case for taxpayers lies in the hope that the Supreme Court goes beyond the administrative law issue and expresses a pro-taxpayer view as to the merits. But this tax issue is not presented.
Rather, the cert petition raises a procedural administrative law issue. It works for the taxpayer only if the government changed its tune. But to the contrary, the government has been singing the same tune for two decades or more.
The government did not surprise taxpayers and tax advisers with never-before-seen interpretations of the arm’s length standard. The government consistently explained that evidence of allegedly comparable transactions is not determinative of an arm’s-length result in this context. It consistently referred to the commensurate-with-income statutory language and legislative intent in support of its position. The government has been faithful to its argument and explanation since before the 2003 amendment and continuing through every stage of this litigation. There has been no surprise or change of course. Rather, this case involves the government making the same argument and explanation, over and over again.
As I have written (I would be happy to supply citations to anyone who wants them), this case is not just about arcane transfer pricing but is about whether Congress writes the law or the Treasury, economists and self-appointed tax elites can overrule Congress. The ubiquity of the arm’s length principle as an essential standard in tax law and the words that Congress used in section 482 should be treated as saying that what matters is what unrelated parties in similar situations do. Period. Full stop. There just is no room for Chevron deference.
This case poses starkly the question of whether that really is the end of it or whether there is scope for the government (or a taxpayer?) to deviate from that for newly conjured reasons, for example, that (i) related parties may have better information than unrelated parties, (ii) a parent may have control over a subsidiary that is absent with unrelated parties (though there may be other commercial pressures giving one more bargaining power than the other) or (iii) for other reasons, including practical convenience and not wanting to get entangled, unrelated parties steadfastly behave in ways that confound theoretical economists.
In short, we can all agree that, on the one hand, parties should allocate all costs, including stock compensation, but, on the other hand, unrelated parties never do that. So can Treasury write a regulation that says that related parties (which includes parties that we might think of as unrelated, given the breadth of “owned or controlled directly or indirectly by the same interests” in section 482) must allocate stock compensation?
We’ve seen movies like this before. For centuries, mathematicians and economists have expressed compound time value of money (interest and discount) with exponential integrals or summations. But in the 1950s and 1960s, before financial calculators became available, a lot of lending still used simple interest or the rule of 78 approximation. Could a regulation have overridden that and required everyone, or at least related parties, to learn how to use logarithm tables? Even now, the statute defines “adequate stated interest”; could a regulation under section 482 demand that related parties, who may know each other’s credit-related strengths, weaknesses and projections also take those into account? Surely not.
This is a compelling case for Supreme Court review because (i) a great deal of money is at stake for a great many taxpayers, (ii) having a unanimous Tax Court take the taxpayer’s side that a regulation is invalid and a divided Ninth Circuit uphold the government marks this case as important, (iii) there is the equivalent of a circuit split (yes, most of the tech companies are in the Ninth, but that can change), and (iv) this is as good a vehicle as any — finally — to discard Chevron deference as a misguided and misapplied doctrine.
Mr. Kantowitz:
First, I like your post. Well stated, shall I say, attack on Chevron. But I wonder whether a frontal assault on Chevron is what Altera made. I thought its argument was not that Chevron was necessarily bad law but that that applying Chevron was inappropriate when the regulation was procedurally invalid (a 706/State Farm) argument I know some blend the two, but I think that they are separate and that procedural invalidity (plus error on the § 482 result) was the heart of Altera’s case, hoping that it can reverse the Ninth Circuit even if the Supreme Court does not view this case as a good vehicle to address Chevron deference. (In this regard, the Court in Kisor v. Willkie just recently rejected the opportunity to eliminate the cousin of Chevron in Auer deference.)
Disclosure, I am not in the ant-Chevron camp, but I do note Professor Pierce’s recent article: Richard J. Pierce, The Combination of Chevron and Political Polarity Will Have Awful Effects, GWU Law School Public Law Research Paper No. 2020-45 (2020)
https://scholarship.law.gwu.edu/cgi/viewcontent.cgi?article=2753&context=faculty_publications
I don’t know enough about Supreme Court procedure, so I will just ask the question that comes to my mind.
On June 1, Altera filed its Reply brief. On June 2, the case was distributed for conference on June 18. As I understand the procedure (from a SCOTUSBlog posting titled Supreme Court procedure):
Once all of the cert stage briefs — the cert petition, the BIO [Brief in Opposition] (if any), the reply brief (if any) and the amicus briefs (if any) — are filed, they are distributed to the justices’ chambers. Seven of the current justices participate in the cert pool, which is a labor-saving device in which a cert petition is first reviewed by one law clerk in one of the seven chambers. That clerk prepares a memorandum about the case that includes an initial recommendation as to whether the court should review the case; the memorandum is circulated to all seven chambers, where it is reviewed by the clerks and possibly the justices there. Justices Samuel Alito and Neil Gorsuch do not participate in the cert pool. Instead, their law clerks review the incoming cert petitions on their own and make recommendations directly to their respective justices.
Based on these reviews, the justices decide to add Lyon v. Animal House Zoo to the discuss list, a short list of cases they plan to talk about at their next private meeting, or conference. (If no justice had asked to add Lyon to the discuss list, it would have been put on the “dead list,” and certiorari would automatically have been denied without the justices having ever discussed the case or voted on it.) At least four justices vote to grant review in Lyon, and the court announces this decision as part of an order list, which will generally be released on the Monday morning after the conference.
JAT Comment: So, because it is on the conference list, presumably there is some interest. But what intrigues me is that there was only one day before the case was put on the conference list for just 16 days off, which I would not think would have allowed time for the cert pool or review process noted above. Perhaps no inference can be drawn from that. Or perhaps, that quick processing would permit an inference that there is already some interest in granting certiorari.
Any thoughts?