Reflections on the Ninth Circuit’s Decision in Amazon.com

Although some time has passed (and we’ve fallen short of our hope here to get something up “soon”), we nevertheless wanted to post some thoughts on the Ninth Circuit’s unanimous affirmance of the taxpayer’s victory in the Tax Court in Amazon.com while also revisiting some of the topics that we covered here after oral argument. The panel’s opinion is brief, but it touched on several important aspects of the law under section 482.

Finding Ambiguity in the Regulatory Definition of “Intangible”

As you’ll recall, the primary dispute on appeal was whether the regulatory definition of “intangible” under Treas. Reg. § 1.482-4(b) included residual business assets like goodwill, going concern, and the amorphous notions of “growth options” and “culture of continuous innovation” that the government theorized the taxpayer made available in its cost-sharing agreement.

The taxpayer argued that the definition excluded such residual business assets because it did not expressly list them in any of the six subparagraphs of the definition. And the taxpayer argued that the 28 specified items in the regulation all “can be sold independently” from the business while the residual business assets cannot, invoking the statutory-interpretation canon of ejusdem generis to reason that the regulation therefore excludes residual business assets (which cannot be sold without a sale of the entire business).

The government argued that those residual business assets fell under the sixth subparagraph of Treas. Reg. § 1.482-4(b), which provides that “intangible” includes “other similar items” and that “an item is considered similar … if it derives its value not from its physical attributes but from its intellectual content or other intangible properties.” Since residual business assets do not derive their value from physical attributes but from other intangible properties, the government reasoned, they must be included here. And the government went on to argue that residual business assets must be compensable because otherwise, taxpayers could have transferred assets of value to a cost-sharing arrangement without compensation, which the government asserts would have violated the arm’s-length principle that undergirds section 482 and its regulations.

The panel held that although the taxpayer’s “focus on the commonality of the 28 specified items has some force,” that argument did not carry the day. The panel drew this conclusion from the plain language in the sixth subparagraph: that paragraph does not state that the commonality is that each item “can be sold independently” but rather states that each item “derives its value … from its intellectual content or other intangible properties.” The panel reasoned that the regulation therefore “leaves open the possibility of a non-listed item being included in the definition even if it doesn’t share the attribute of being separately transferable.” The panel thus echoed the concern that Judge Fletcher raised at oral argument—that while the 28 specified items share the commonality of being independently transferable, that commonality is “not the one the text [describing what constitutes a ‘similar item’ under the regulation] gives me.”

Resolving the Ambiguity by Looking to the Drafting History of the Regulation

After concluding that the regulatory definition “is susceptible to, but does not compel, an interpretation that embraces residual-business assets,” the panel looked to the overall regulatory scheme but held that to be inconclusive. The panel then turned to the drafting history of Treas. Reg. § 1.482-4(b) and found it to show that Treasury’s 1994 final regulations (which amended the sixth subparagraph of Treas. Reg. § 1.482-4(b)) both (1) exclude residual business assets from the definition of intangible and (2) provide reason to think that the definition of “intangible” includes only independently transferable assets.

The panel described how when Treasury issued temporary and proposed regulations in 1993, it asked “whether the definition of intangible property … should be expanded to include … goodwill or going concern value.” The panel concluded that since Treasury asked whether the definition should be “expanded” to include residual business assets, those assets were not included in the then-existing definition. And when Treasury issued final regulations in 1994 without expressly enumerating goodwill or going concern value in that definition, Treasury stated that its final rule “merely ‘clarified’ when an item would be deemed similar to the 28 items listed in the definition.” The panel concluded that by its own admission, Treasury would have needed to “expand” the definition to include residual business assets but instead opted to merely “clarif[y]” its definition, and therefore Treasury did not intend for the 1994 final regulations to include residual business assets.

Moreover, Treasury’s 1993 regulations limited the universe of compensable intangibles to “any commercially transferable interest.” But when it issued final regulations in 1994, Treasury dropped the “commercially transferable interest” language because “it was superfluous: if the property was not commercially transferable, then it could not have been transferred in a controlled transaction.” The panel concluded that the transfer-pricing regulations thus “contemplate a situation in which particular assets are transferred from one entity to another.” Since Treasury stated that it would have been “superfluous” to expressly state that the definition of “intangible” includes only commercially transferable assets, the panel held that the regulatory history “strongly supports Amazon’s position that Treasury limited the definition of ‘intangible’ to independently transferable assets.”

That the panel found this history dispositive comes as no surprise. We observed in our prior post that both Judge Callahan and Judge Christen recounted the regulatory history, with Judge Callahan questioning government counsel about whether Treasury ever expressed an intent to expand the definition to include residual business assets.

What Is the Rationale for Including Only Independently Transferable Assets in the Definition of Intangible?

Although the panel’s explanation for how it concluded that the definition of intangibles includes only independently transferable assets is explicit, the rationale for why Treasury would include only independently transferable assets is markedly subtler. It’s possible, however, to cobble together an explanation from other statements in the decision.

The first clue is when the panel looked to the genesis of the cost-sharing regulations, where Treasury identified “intangibles as being the product of R&D efforts.” In that sense, the panel reasoned, the “regulations seem to exclude” residual business assets, “which ‘are generated by earning income, not by incurring deductions.’” This distinction is clear enough—businesses incur expenses in undertaking R&D efforts, while goodwill and going concern value are byproducts of a well-run and successful business.

Why does this distinction matter? One answer lies in the legislative history and policy underpinnings for the statutory definition of “intangibles.” As the panel observed in a footnote, the “Senate Report states that the Committee viewed the bill as combatting the practice of transferring intangibles ‘created, developed or acquired in the United States’ to foreign entities to generate income tax free.” Which is to say that one animating concern in defining intangibles was to prevent U.S. entities from developing intangibles domestically but then transferring those intangibles to foreign affiliates whose income is not subject to U.S. tax.

And the reason why this is a concern for intangibles that result from R&D and other independently transferrable assets but not goodwill or going-concern value should be apparent: while the former involve expenditures that are deductible against U.S. income (but where the asset transfer will prevent the income from being taxed in the U.S.), the latter involve assets that exist only if the taxpayer has generated business income in the U.S. in the first place. In other words, the definition of “intangible” was initially meant to prevent taxpayers from incurring expenses to create intangibles in the U.S. and deducting those expenses against U.S. income, but then turning around and transferring those intangibles to foreign affiliates that may not owe U.S. tax on the income resulting from those intangibles.

Distinguishing the Definition of “Intangibles” from the Value of Those Intangibles

On brief and at oral argument, the government repeatedly cited deposition testimony by one of the taxpayer’s experts in which that expert admitted that parties at arm’s length would pay for residual business assets. The government tried to leverage that admission to argue that the arm’s-length standard itself means that residual business assets are compensable because “it is undisputed that a company entering into the same transaction under the same circumstances with an unrelated party would have required compensation.”

The panel addressed that argument in a footnote, holding that the government’s argument “misses the mark.” The panel explained that while the arm’s-length standard “governs the valuation of intangibles; it doesn’t answer whether an item is an intangible.” This is a decisive response to the government’s arguments; it cannot be the case that any value associated with a business falls under the definition of “intangible.”

Putting Footnote 1 in Context

One aspect of the panel’s decision that is certain to receive attention in future transfer-pricing disputes is the discussion in the first footnote. In that footnote, the panel described the 2009 changes to the cost-sharing regulations as “broadening the scope of contributions for which compensation must be made” and explained that the TCJA “amended the definition of ‘intangible property’” in section 936(h)(3)(b). The footnote then stated that “[i]f this case were governed by the 2009 regulations or by the 2017 statutory amendment, there is no doubt the Commissioner’s position would be correct.”

There are a few things worth noting about this footnote, the first of which is arguably the most important: It’s dicta. The question of whether residual assets are compensable under the 2009 cost-sharing regulations or the 2017 statutory amendment was not before the court. And since the panel says nearly nothing in the footnote about the language in those 2009 regulations or the TCJA amendment, there is no reason to think that the panel gave material consideration to whether the outcome in the case would have been different under either.

Second, given some of the panel’s other statements, it is not so clear that “there is no doubt the Commissioner’s position would be correct” for years after the 2009 changes to the regulations. Recall that the government’s case was predicated on charging a higher buy-in for the purported transfer of “growth options” and a “culture of continuous innovation” to the cost-sharing arrangement. Although the 2009 regulatory amendments changed the cost-sharing regulations and TCJA amended the statutory definition for “intangibles,” there was no change to the definition under Treas. Reg. § 1.482-4(b). That regulation provides that an item not otherwise specified is an intangible only if it “has substantial value independent of the services of any individual.” The panel itself expressed serious doubt about whether the purported intangibles in this case met that requirement, stating that “residual business assets, such as ‘growth options’ and a ‘culture of innovation,’ are amorphous, and it’s not self-evident whether such assets have ‘substantial value independent of the services of any individual.’”

Finally, the TCJA amendments would affect the result in the case only if the purported “growth options” and “culture of continuous innovation” are items of property that fall under the new statutory category of “goodwill, going concern value, or workforce in place” or constitute an item “the value or potential value of which is not attributable to … the services of any individual.” But there is reason to think that the government is itself unconvinced that “growth options” and “culture of continuous innovation” fall under that new statutory category. At oral argument, the panel asked government counsel why, if Treasury meant to expand the definition of “intangible” with its 1994 changes, Treasury didn’t just add “goodwill and going concern” to the list. In response, government counsel argued that merely adding “goodwill and going concern” to the list would not have solved anything because then taxpayers would just fight about whether particular assets fell under those additions to the list. The government cannot coherently maintain both that (1) Treasury could not have achieved the Commissioner’s desired result in this case by expanding the regulatory list to include goodwill and going concern in 1994 and (2) Congress’s addition of goodwill and going concern to the list in TCJA would change the result in this case. Even though Congress meant to change the result in cases like this, government counsel’s argument suggests an absence of faith that expanding the list succeeds in changing the result.

Whether the Commissioner’s Litigating Position Warrants Auer Deference

The government also argued that under Auer, the Tax Court should have deferred to the Commissioner’s interpretation of Treas. Reg. § 1.482-4(b) as including residual business assets. The panel rejected this argument for two reasons.

First, the panel found that pursuant to the Supreme Court’s decision in Kisor, the Commissioner’s interpretation warrants Auer deference only if the regulation is “genuinely ambiguous.” The panel states that Auer thus implicates a higher standard for ambiguity. Under that standard, a regulation is “genuinely ambiguous” for Auer purposes only if ambiguity remains once the court has exhausted the traditional tools of construction, which requires it to consider “the text, structure, history, and purpose of a regulation.” The panel concluded that the text of Treas. Reg. § 1.482-4(b), its place in the transfer-pricing regulations, and its rulemaking history “leave little room for the Commissioner’s proffered meaning.” The panel’s holding is therefore that the definition of “intangible” under Treas. Reg. § 1.482-4(b) is ambiguous but not genuinely ambiguous.

Second, the panel also rejected the government’s deference argument because the government first advanced an interpretation of Treas. Reg. § 1.482-4(b) in the litigation that had never appeared in the drafting history of the regulations or anywhere else. As we remarked earlier, reliance concerns loomed large at oral argument, and especially in Judge Callahan’s questions. The panel’s conclusion that “Amazon and other taxpayers were not given fair warning of the Commissioner’s current interpretation of the regulatory definition of an ‘intangible’” was thus foreseeable from oral argument.

Whether the Cost-Sharing Regulations Provide a Safe Harbor

As we observed in our prior post, at oral argument the government disavowed the notion that the Treasury Regulations create a safe-harbor for cost-sharing arrangements. We surmised that whether the Ninth Circuit agreed with the government on this point might be pivotal in the outcome.

While the extent to which that issue factored into the panel’s decision is not evident, the panel’s decision is entirely consistent with the notion that cost sharing operates as a safe harbor. First, the panel described cost sharing as “an alternative to licensing … under which [the parties to the cost-sharing arrangement] become co-owners of intangibles as a result of the entities’ joint R&D efforts.” If the cost-sharing arrangement qualifies, then it “provides the taxpayer the benefit of certainty because … new intangibles need not be valued as they are developed.” But the panel explained that the certainty comes at a price—the R&D payments by the foreign cost-sharing participants “serve to reduce the deductions the [domestic] taxpayer can take for the R&D costs (thereby increasing tax liability).” What the panel thus described operates like a safe harbor—taxpayers that ensure that their cost-sharing arrangements qualify and sacrifice some deductions for intangible development costs can gain certainty that their intangibles need not be re-valued. The panel’s decision will hamper any future IRS arguments that cost-sharing does not operate like a safe harbor.

Procedural Status

The government did not file a petition for rehearing in the case; the mandate issued on October 8. There is still time for the government to file a petition for certiorari; it is due November 14.

Ninth Circuit Affirms Tax Court in Amazon.com

In a unanimous opinion issued today, the Ninth Circuit affirmed the taxpayer’s victory in the Tax Court in Amazon.com. We previously covered the case and oral argument here. We will take some time to digest the opinion and post on its finer points soon. In the meantime, one key sentence in the opinion is worth noting because it appears to capture the thrust of the Ninth Circuit’s decision about the disputed scope of the relevant regulatory definition for the term “intangible”: “Although the language of the definition is ambiguous, the drafting history of the regulations shows that ‘intangible’ was understood to be limited to independently transferable assets.”

Amazon.com Ninth Circuit Opinion

Petition for Rehearing En Banc Filed in Altera

As most expected, Altera filed a petition for rehearing en banc after the reconstituted three-judge panel decided to reverse the Tax Court’s invalidation of Treasury’s cost-sharing regulations. (A link to the petition is below.) As we explained previously, those regulations have been the subject of much controversy over the last two decades, and the success that Xilinx had with its petition for rehearing several years ago made it likely that Altera would ask for rehearing.

The petition picks up on one of the themes we discussed in our most recent post here. The taxpayer takes aim at the majority’s conclusion that the “commensurate with income” language added to section 482 in 1986 is relevant in the cost-sharing context. The taxpayer argues that language was aimed at addressing a different issue from the one before the court here—“how to value transfers of existing intangible property from one related entity to another” and not “intangible property yet to be created.”

The taxpayer makes four or five (the petition combines arguments (3) and (4) below) arguments for why its petition should be granted:

(1) The decision upsets settled principles about the application of the arm’s-length standard because the majority permitted Treasury to “cast aside the settled arm’s-length standard” for “a new standard” that is “purely internal.”

(2) The decision “validates bad rulemaking” because, contrary to the majority’s account of the regulation’s history, “[n]o one involved in the rulemaking thought the IRS was interpreting ‘commensurate with income’ to justify a new standard that did not depend on empirical evidence.” And under the law in Chenery, the court must assess the “‘propriety of [the agency’s] action solely by the grounds invoked by the [agency]’ in the administrative record.”

(3) The decision is irreconcilable with the Ninth Circuit’s decision in Xilinx, which held that parties would not share in employee stock option costs at arm’s length.

(4) The decision “threatens the uniform application of the tax law” because, under the Golsen rule, the Tax Court will continue to apply its unanimous decision declaring the regulation invalid to cases arising anywhere outside the Ninth Circuit.

(5) As evidenced by the glut of amicus briefs, the treatment of employee stock options in cost-sharing arrangements is “exceptionally important.”

It is likely that additional amicus briefs will be filed in support of the rehearing petition. And given the prominence of the issue, we anticipate that the court will order the government to file a response to the petition. We will report on further developments as warranted.

Altera Petition for Rehearing En Banc July 2019

Observations on Changes in the Ninth Circuit’s Second Altera Decision

As we posted earlier here (with a link to the new decision), the Ninth Circuit issued a new decision in Altera after replacing the late Judge Reinhardt with Judge Graber on the panel. But the result was the same as the withdrawn July 2018 decision—the Ninth Circuit upheld the validity of Treasury’s cost-sharing regulation that requires taxpayers to include the cost of employee stock options under qualifying cost sharing arrangements (QCSAs). Today, we present some observations after comparing the majority and dissent in the new decision with those in the Ninth Circuit’s withdrawn decision.

In the new decision, Judge Thomas recycled much of the language and logic from his withdrawn opinion, and Judge O’Malley reused much of her original dissent. Although they are few, some changes in the two opinions are interesting and notable. Overall, the changes serve to sharpen the disagreements between the parties (and the disagreements between the majority and dissent) in ways that will focus the discussion in the likely event of a rehearing en banc petition (or possible petition for certiorari). We focus here on two aspects of the changes.

Was There a Transfer of Intangibles that Implicated the Commensurate-With-Income Language in the Second Sentence of Section 482?

The majority did not address this issue in its withdrawn opinion, so some background is in order. When Treasury proposed cost-sharing regulations that explicitly required related parties to include employee-stock-option costs in the pool of shared costs, commenters put forward evidence that unrelated parties do not share employee-stock-option costs. But Treasury did not heed those comments and ultimately determined that the arm’s-length standard would be met if the regulations required taxpayers in QCSAs to include the cost of employee stock options in the pool of shared costs, regardless of what a comparability analysis might show about whether unrelated parties share those costs. So in order to uphold the Treasury Regulations, the majority had to conclude that the arm’s-length standard under section 482 does not mandate the use of comparable transactions.

In reaching that conclusion, the majority relied on the history of section 482 and especially on the addition of the second sentence of section 482. That second sentence provides that “[i]n the case of any transfer (or license) of intangible property…, the income with respect to such transfer or license shall be commensurate with the income attributable to the intangible.” The majority held that the Congressional intent behind the addition of that language in 1986 is what made it reasonable for Treasury to conclude that it was permitted “to dispense with a comparable transaction analysis in the absence of actual comparable transactions.”

Was there a “transfer (or license) of intangible property” such that Treasury could invoke the commensurate-with-income language to justify dispensing with a comparable-transaction analysis? There was indeed a transfer of intangibles at the outset of the QCSA in this case; Altera transferred intangible property to its QCSA with its foreign subsidiary. But that initial transfer is unrelated to the disputed employee-stock-option costs. Those stock-option costs relate to subsequently developed intangibles, not to the value of the pre-existing intangibles that Altera initially contributed. Then-applicable Treasury Regulation § 1.482-7A(g)(2) required a buy-in payment only for “pre-existing intangibles.”

It is, then, wholly unclear how those employee-stock-option costs for as-yet undeveloped intangibles constitute a “transfer (or license) of intangible property” such that the commensurate-with-income language is implicated at all. The taxpayer argued—both in its initial and supplemental briefing—that by its own terms, the commensurate-with-income language is not implicated once a QCSA is in place and therefore that language is irrelevant to the dispute. The taxpayer reasoned that “no ‘transfer (or license)’ occurs when entities develop intangibles jointly” because jointly developed intangibles “are not transferred or licensed between the parties, but rather are owned upon their creation by each participant.”

The majority said little about this issue in its withdrawn opinion and did not address the taxpayer’s argument. In its new opinion, the majority addressed the argument and declared itself “unpersuaded.” It held that “[w]hen parties enter into a QCSA, they are transferring future distribution rights to intangibles, albeit intangibles that are yet to be developed.” The majority does not, however, explain how “intangibles that are yet to be developed” are “pre-existing intangible property” under Treas. Reg. § 1.482-7A(g)(2) or, perhaps more importantly, how such undeveloped future intangibles can constitute intangible assets at all under the words of the statute. Instead, the majority leaned on the language in the second sentence of section 482 providing that the commensurate-with-income standard applies to “any” transfer of intangible property, holding that “that phrasing is as broad as possible, and it cannot reasonably be read to exclude the transfers of expected intangible property.”

The dissent seized on whether there was any transfer of intangibles that implicated the commensurate-with-income language. The dissent stated that “[t]he plain text of the statute limits the application of the commensurate with income standard to only transfers or licenses of intangible property.” And it observed that there is a contradiction between the majority’s conclusion that QCSAs “constitute transfers of already existing property” and Treasury’s own characterization (in the very preamble of the disputed regulations) of QCSAs “as arrangements ‘for the development of high-profit intangibles,’” inferring that parties cannot transfer something that is not yet developed. The dissent concluded that Treasury’s failure to make “a finding that QCSAs constitute transfers of intangible property” should be, as a matter of review under the APA, fatal to Treasury’s cost-sharing regulations.

Has the Arm’s-Length Standard Historically Required a Comparable-Transaction Analysis?

The majority made several changes from its withdrawn opinion in its account of the history of the arm’s-length standard under section 482. The changes appear to be aimed at bolstering the majority’s conclusion that Treasury was justified in concluding that the arm’s-length standard does not necessitate a comparable-transaction analysis. These changes take a couple of forms.

First, the changes add justifications for the proposition that the arm’s-length standard has not always mandated a comparable-transaction analysis in all cases. The withdrawn opinion included some evidence for this historical account, recounting the Tax Court’s Seminole Flavor decision from 1945 (where the Tax Court “rejected a strict application of the arm’s length standard in favor of an inquiry into whether the allocation…was ‘fair and reasonable’”). In its new opinion, the majority added the observation that Treasury provided for an unspecified fourth method for pricing intangibles in its 1968 regulations. And the majority refined its discussion of the Ninth Circuit’s 1962 decision in Frank, still quoting the language from that opinion denying that “‘arm’s length bargaining’ is the sole criterion for applying” section 482 (as it did in the withdrawn opinion) but then adding the new assertion that the “central point” of Frank is that “the arm’s length standard based on comparable transactions was not the sole basis” for reallocating costs and income under section 482. (As the dissent pointed out, a later Ninth Circuit decision limited the holding in Frank to the complex circumstances in that case and noted that the parties in Frank had stipulated to apply a standard other than the arm’s-length standard.)

Second, the majority made changes to add a series of new and more sweeping assertions that under section 482, Treasury has always had the leeway to dispense with comparable-transaction analyses in deriving the correct arm’s-length price. For instance, as a rejoinder to the taxpayer’s argument that the arm’s-length standard requires a comparable-transaction analysis, the majority wrote that “historically the definition of the arm’s length standard has been a more fluid one” and that “courts for more than half a century have held that a comparable transaction analysis was not the exclusive methodology to be employed under the statute.” It added similarly broad historical statements elsewhere in the new opinion, all in service of concluding that Treasury acted reasonably in dispensing with a comparable-transaction analysis in its cost-sharing regulations: “[a]s demonstrated by nearly a century of interpreting § 482 and its precursor, the arm’s length standard is not necessarily confined to one methodology” (p. 33); “the arm’s length standard has historically been understood as more fluid than Altera suggests” (p. 41); and “[g]iven the long history of the application of other methods…Treasury’s understanding of its power to use methodologies other than a pure transactional comparability analysis was reasonable” (p. 49).

In one of the subtler but more interesting changes from the withdrawn opinion, the majority’s new opinion removed a single word. In its withdrawn opinion, the majority said that Treasury’s 1988 White Paper “signaled a dramatic shift in the interpretation of the arm’s length standard” by advancing the “basic arm’s length return method…that would apply only in the absence of comparable transactions….” (emphasis added). But consistent with its conclusion in the new opinion that “for most of the twentieth century the arm’s length standard explicitly permitted the use of flexible methodology,” the majority appears to have concluded that shift in the White Paper was not so “dramatic,” dropping that word altogether in its new opinion. (In this vein, the majority also removed its assertion in the withdrawn opinion that “[t]he novelty of the 1968 regulations was their focus on comparability.”)

The new dissenting opinion disputed the majority’s historical account, stating that the first sentence of section 482 “has always been viewed as requiring an arm’s length standard” and that before the 1986 amendment, the Ninth Circuit “believed that an arm’s length standard based on comparable transactions was the sole basis for allocating costs and income under the statute in all but the narrow circumstances outlined in Frank.” And the dissent observed that even with the 1986 addition of the commensurate-with-income language, “Congress left the first sentence of § 482—the sentence that undisputedly incorporates the arm’s length standard—intact,” thus requiring a comparable-transaction analysis everywhere that comparable transactions can be found. The dissent pointed out that the White Paper clarified that this was true “even in the context of transfers or licenses of intangible property,” quoting Treasury’s own statement in the White Paper that in that context the “‘intangible income must be allocated on the basis of comparable transactions if comparables exist.’”

Briefing Complete in Kisor

The petitioner has now filed his reply brief in Kisor, and the case is fully briefed in preparation for the oral argument later this week on March 27.  Given the government’s partial retreat from defending Auer deference (see our prior post here), which the petitioner describes as a “sharp retreat,” the reply brief responds to two different briefs.  First, it responds directly to an amicus brief by a group of law professors (linked in our prior post) that put forth a full-throated defense of Auer deference.  Second, it acknowledges that the government’s “Auer-light” position is “preferable to existing Auer deference,” but it still rejects that position and argues that complete overruling of Auer is the correct approach.  The reply brief concludes that the existing principles of Skidmore deference satisfactorily address the policy goals described in the government’s brief without improperly permitting “an agency to exert its expertise in binding fashion without any participation by the regulated public.”

Kisor – Petitioner Reply Brief

 

Treasury and IRS Issue Joint Policy Statement on the Tax Regulatory Process

Earlier this week, the Treasury Department issued a policy statement on the tax regulatory process.  A significant section of the statement describes the approach that will be taken in Tax Court litigation to arguments based on judicial deference to regulations.  Treasury states that it will not claim Auer deference in such litigation to interpretations set forth in subregulatory guidance, such as revenue rulings, nor will it claim Chevron deference to such interpretations.  That apparent abandonment of Auer deference arguments goes beyond the position the Justice Department has taken in the Supreme Court in the Kisor case, where the government has argued for a significant narrowing of Auer but stops short of stating that it should be overruled.  See our reports on Kisor here.  Note that the Treasury Department statement by its terms applies only to Tax Court litigation, which is handled by the IRS, not to refund suits where the Justice Department handles the litigation.

The Treasury Department statement also addresses topics other than judicial deference, including: when to issue subregulatory guidance; legislative vs. interpretative regulations; and temporary regulations.  For a fuller analysis of the Treasury Department statement, please read this Miller & Chevalier Tax Alert.

Government Brief Filed in Kisor

The government was faced with something of a dilemma in filing its response brief in the Kisor case addressing the level of deference owed to an agency’s interpretation of its own regulation. See our prior reports here. On the one hand, the government was defending the agency action in this case and the decision below, which rested on paying Auer deference to the agency’s interpretation. On the other hand, conservative legal theorists have long been critical of Auer deference, following Justice Scalia’s lead, and the views of the political appointees in this administration about Auer likely range from unenthused to hostile. But on the third hand, the government’s institutional interests would generally be better served by a strong principle of Auer deference, since that would make challenges to agency action more difficult.

The government’s brief attempts to juggle these conflicting imperatives, and the result is a bit schizophrenic. The bottom line is that the government argues that Auer should not be overruled, but that its applicability should be substantially narrowed. In the end, the government does not rely on Auer to defend the agency action in this case, but instead argues that the regulation is clear on its face without the need to consider the agency interpretation at all.  (Although outgunned by the cascade of amicus briefs filed in support of the petitioner, two amicus briefs were filed on the government’s side, including one by a group of administrative law professors (linked below) who argue that Auer “is sound and should be maintained.”)

The first, and longest, section of the government’s brief is a full-fledged assault on the doctrine of Auer deference. The government contends that the doctrine: (1) is not well grounded historically; (2) is not supported by any consistent rationale; (3) is in tension with the APA’s distinction between interpretive and legislative rules; and (4) can have harmful practical consequences by discouraging agency resort to notice-and-comment rulemaking. Notably, the government states that the reasons that support Chevron deference do not apply to Auer, and thus the brief does not signal that the current administration will argue against Chevron deference in a future case.

The government argues, however, that Auer should not be overruled because of stare decisis considerations, including that doing so “would upset significant private reliance interests” because it allegedly “could call into question” earlier decisions that rested on Auer deference. In contrast to the opening part of its brief, this section praises Auer deference where it is limited to “its core applications,” such as “when the agency announces its interpretation in advance in a widely available guidance document.” The government states that the task of choosing among reasonable interpretations is more appropriately performed by administrators than by judges, that Auer deference would promote national uniformity, that it recognizes the technical expertise of agencies, and that it fosters regulatory certainty and predictability—in contrast to a system “in which the meaning of a regulation must be determined de novo in every judicial proceeding.” In addition, the government disagrees with the petitioner’s argument that Auer deference poses a separation-of-powers problem, stating that an agency’s actions in making rules and conducting adjudications are both exercises of “executive power.”

Accordingly, the government proposes “significant limits” on the doctrine that will thread the needle, neither overruling Auer nor further entrenching it. First, the government states that deference should not be paid to an agency interpretation that is “unreasonable,” describing this seemingly benign limitation as a “rigorous predicate.” If the agency interpretation is judged to be within the range of reasonable readings of the regulation, then the government argues that deference is appropriate “only if the interpretation was issued with fair notice to regulated parties; is not inconsistent with the agency’s prior views; rests on the agency’s expertise; and represents the agency’s considered view, as distinct from the views of mere field officials or other low-level employees.”

It is hard to say at this point what the Court will do with the various permutations that have been presented to it for moving forward, but it appears that Auer deference in its current form stands on very shaky ground.

Oral argument is scheduled for March 27.

Kisor – Government Response Brief

Kisor – Amicus Brief by Administrative Law Scholars in Support of Auer

Government Brief Filed in MassMutual

The government has filed its opening brief in MassMutual contesting the Court of Federal Claims’ conclusion that the taxpayer could accrue the amount of certain policyholder dividends in the year before they were paid.  See our prior post on this case and the New York Life case here.  The government’s brief raises three distinct objections to the decision.

The primary argument is that the liability to pay the dividends was not “fixed” under the all-events test.  The government contends that no individual obligation was fixed at the close of the year, even if all the premiums had been paid, because the dividend would not be paid unless the policy remained in force on the anniversary date.  This is the same argument that was accepted by the Second Circuit in New York Life, and the government’s brief here argues that the cases are indistinguishable (asserting that the Second Circuit’s effort to distinguish them was based on a misperception of the facts in MassMutual).

The brief argues that the case “clearly fits the General Dynamics fact pattern,” which it describes as one where the “potential obligee has taken some action that renders him preliminarily eligible to receive the payment, subject only to some other condition that is within his exclusive control” – here, “forgoing the right to surrender the policy for its cash value prior to the next anniversary date.”  It rejects the proposition argued by the taxpayer that this alleged final condition is not a genuine “event,” but rather just a continuation of the status quo.  The government points to a comment in the Restatement (Second) of Contracts stating that “a duty may be conditioned upon the failure of something to happen . . ., and in that case its failure to happen is the event” that constitutes a condition precedent.  And it rejects the contrary suggestion in Burnham Corp. v. Commissioner, 878 F.2d 86 (2d Cir. 1989), as misguided.  Finally, the brief argues that the taxpayer’s all-events-test interpretation proves too much because its logical implication is that the amount of the dividend could be accrued even if the company had not passed a board resolution in the taxable year guaranteeing an aggregate dividend – a position that the taxpayer has not argued.

Second, the government argues that the dividend guarantees did not even give rise to an obligation, fixed or otherwise, because they were not communicated to the persons who were to benefit from them.  Thus, the government argues, the taxpayer could have walked away from the guarantees at any time.  In addition, the government argues, the guarantees were not a meaningful “substantive undertaking” because, based on the historical data, the guaranteed payments were “already virtually certain to occur in the ordinary course of the companies’ business operations, independent of any ‘guarantee’ to that effect.”  There is some degree of irony in this argument; on its face, certainty that the amounts will be paid would appear to be an argument in favor of accrual, not against it.  But the certainty of which the government speaks refers to the aggregate amount of payment; it is not a concession with respect to an individual obligation being fixed.

Third, the government contests the Court of Federal Claims’ holding that the dividends fell within the “recurring item” exception.  The government’s primary point here is that this determination turns on the meaning of “rebate, refund, or similar payment” in Treas. Reg. § 1.461-4(g)(3), and therefore the court should have deferred to the IRS’s interpretation of that regulation – even if that interpretation did not conclusively emerge until this litigation and is at odds with some earlier internal guidance on the regulation’s meaning.  The general principle of so-called Auer or Seminole Rock deference to an agency’s interpretation of its own regulations has come under fire recently, with Justice Scalia stating that it should be abandoned and Chief Justice Roberts and Justice Alito indicating that they are at least open to reconsidering it.  See Decker v. Northwest Environmental Defense Center, No. 11-338 (Mar. 20, 2013).  So it will be interesting to see how the Federal Circuit responds to this argument, which presents a relatively weak case for deference because the claimed agency interpretation is just based on its litigation position.

The taxpayer’s brief is due April 4.

MassMutual – Gov’t Opening Brief

D.C. Circuit Holds in Loving that IRS Lacks Authority to Regulate Tax-Return Preparers

February 12, 2014 by  
Filed under Loving, Regulatory Deference

Yesterday, the D.C. Circuit unanimously held in Loving v. IRS, that the IRS lacks statutory authority to regulate tax-return preparers.  See our previous coverage here.  In its February 11 decision, the court characterized the IRS’s interpretation as “atextual and ahistorical,” and, more humorously, as a large elephant trying to emerge from a small mousehole.

In 2011, the IRS for the first time attempted to regulate tax-return preparers, issuing regulations requiring that paid tax-return preparers pass an initial certification, pay annual fees, and complete at least 15 hours of continuing education courses each year.  The IRS estimated that the regulations would apply to between 600,000 and 700,000 tax-return preparers.  Before 2011, the IRS had never taken the position that it had the authority to regulate tax-return preparers.  According to the D.C. Circuit panel (Kavanaugh, Sentelle, and Williams): “In light of the text, history, structure, and context of the statute, it becomes apparent that the IRS never before adopted its current interpretation for a reason:  It is incorrect.”

The IRS claimed that 31 U.S.C. § 330 provided statutory authority for the regulations.  That statute authorizes the IRS to “regulate the practice of representatives of persons before the Department of Treasury.”  31 U.S.C. § 330(a)(1).  The D.C. Circuit cited the familiar two-step Chevron standard of review:  (1) is the statute ambiguous, and (2) if so, is the agency’s interpretation reasonable.   The court of appeals concluded that the “IRS’s interpretation fails at Chevron step 1 because it is foreclosed by the statute,” and, in any event, “would also fail at Chevron step 2 because it is unreasonable in light of the statute’s text, history, structure, and context.”  The court of appeals cited six reasons foreclosing the IRS’s interpretation of the statute:

First, the D.C. Circuit concluded that the term “representative” generally is understood to refer to an agent with authority to bind others.  “Put simply, tax-return preparers are not agents.  They do not possess legal authority to act on the taxpayer’s behalf.”  Second, the preparation of a tax return does not constitute “practice . . . before the Department of the Treasury.”  “Practice before” an agency generally implies an investigation, adversarial hearing, or other adjudicative proceeding.  Moreover, a related section of the statute allows the Secretary of the Treasury to require that a representative admitted to practice before the agency demonstrate four qualities, one of which is “competency to advise and assist persons in presenting their cases.”  31 U.S.C. § 330(a)(2).   Filing a tax return is not understood in ordinary usage to be “presenting a case.”  Third, the original version of the statute, enacted in 1884, referred to “agents, attorneys, or other persons representing claimants before [the] Department.”  The court of appeals concluded that this original language clearly would not encompass tax return preparers.  When the statute was recodified in 1982, the phrase, “agents, attorneys, or other persons representing claimants” was simplified to “representatives of persons,” but the language change expressly was not intended to effect a substantive change.  Fourth, the IRS’s interpretation is inconsistent with the “broader statutory framework,” in which Congress has enacted a number of statutes specifically directed at tax-return preparers and imposing civil penalties.  Those statutes would not have been necessary, the court reasoned, if the IRS had authority to regulate tax-return prepares.  Fifth, if Congress had intended to confer such broad regulatory authority upon the IRS, allowing it to regulate “hundreds of thousands of individuals in the multi-billion dollar tax-preparation industry,” the statute would have been clearer.  Referring to the statutory language, the court of appeals concluded:  “we are confident that the enacting Congress did not intend to grow such a large elephant in such a small mousehole.”  Sixth, the court noted that the IRS in the past had made statements and issued guidance indicating that it did not believe it had authority to regulate tax-return preparers.  The court found it “rather telling that the IRS had never before maintained that it possessed this authority.”

The decision ends with the court of appeals noting that new legislation would be needed to allow the IRS to regulate tax-return preparers.

Given that the membership of the D.C. Circuit has recently expanded to include three additional judges, the government might believe that it is worthwhile to seek rehearing en banc before the full court.  A petition for rehearing would be due on March 28.  If the government does not seek rehearing, a petition for certiorari would be due on May 12.  Whether it pursues the litigation further or not, the government can be expected to seek new legislation that would give the IRS the regulatory authority that the court of appeals refused to find.

Loving – Court of Appeals Opinion

Government Faces Sharp Questioning from D.C. Circuit in Loving

The D.C. Circuit heard oral argument on September 24 in the government’s appeal in Loving from the district court decision enjoining the IRS from enforcing its new registration regime for paid tax return preparers.  The panel consisted of Judges Sentelle, Williams, and Kavanaugh.  The court was active, jumping in with questions in the first minute of the government’s opening presentation.  The court asked several questions of the plaintiffs’ counsel as well, but those questions seemed to evince less skepticism of the advocate’s position.  While it is always hazardous to predict the outcome based on the oral argument, the court of appeals certainly seemed to be leaning towards affirming the district court.

As we have previously discussed, the government’s position relies heavily on Chevron deference to its new tax return preparer regulations.  It argues that the statutory authority to regulate practice before Treasury is sufficiently broad to encompass tax return preparers — specifically, that the term “practice of representatives of persons before the Department of the Treasury” is ambiguous and could reasonably be construed by the regulations to include persons who prepare tax returns.  The relevant language is currently codified at 31 U.S.C. § 330(a)(1), but it dates back to 1884, when Congress responded to complaints about misconduct by claims agents who represented soldiers with claims for lost horses or other military-related compensation from the Treasury.

Just 30 seconds after the argument began, Judge Sentelle stepped in to challenge the premise of government counsel Gil Rothenberg that the Treasury regulations were valid because the statute did not “foreclose” them.  Judge Sentelle maintained that the question instead was whether the statute “empowered” Treasury to regulate in this area, and the case could not be analyzed by assuming that Treasury had unlimited power except to the extent that Congress had explicitly foreclosed it.  Shortly thereafter, Judge Williams questioned the government’s failure, in his view, to provide any support for the notion that the ordinary use of the statutory terms, like “representative” or “practice,” could encompass a tax return preparer who merely helps a taxpayer “fill out a form” that he is obliged to file with the IRS.  Mr. Rothenberg responded that, although there were no cases on point, return preparers do more than “fill out a form” and that the statutory term “representative” cannot be limited to an agency relationship because the term was intended to retain the same meaning as the original 1884 statute, which applied to “agents, attorneys, or other persons representing claimants.”

Judge Sentelle then suggested that the fact that Treasury had not claimed any authority to regulate tax return preparers until now, even though the statute had been on the books for more than a century, cast some doubt on the existence of that authority.  Judge Kavanaugh added that Congress’s enactment of legislation regulating tax return preparers during that period also suggested that Congress did not think that it had delegated that authority to the Treasury Department.  Mr. Rothenberg responded that the administrative process is an “evolving process,” and Treasury was free to “choose” not to regulate for many years and then later to invoke its latent authority to regulate.  Later, he added that the need to regulate the competence of tax return preparers is greater today than it was decades ago when taxpayers could more easily avail themselves of direct assistance from the IRS in filling out their return.  With respect to the legislation, Mr. Rothenberg distinguished laws that impose after-the-fact sanctions on return preparers from the Treasury initiative to impose up-front “admission” requirements.  Judge Sentelle questioned why the regulations are limited to paid return preparers, but do not cover persons who prepare tax returns for free.  Mr. Rothenberg responded that Treasury was tackling the problem one step at a time and reasonably believed that the biggest problem was with unqualified persons marketing their ability to prepare returns.

Judge Kavanaugh then zeroed in on the statutory text, pointing out that section 330 (a)(2)(D) states that Treasury “may . . . require” that “representatives” demonstrate their “competency to advise and assist persons in presenting their cases.”  That language indicates that Congress understood that the “representatives” who could be regulated were persons who would assist in “presenting cases,” not just filling out returns.  Mr. Rothenberg disagreed, arguing that Treasury was not compelled to impose all of the requirements set forth in subsection (a)(2) and that the other three requirements could apply to tax return preparers.  Judge Sentelle expressed some doubt whether that position was consistent with the statutory use of the conjunctive “and” in joining the four subsections of (a)(2).  Judge Williams then suggested that these were four different characteristics of representatives, but that the language of (a)(2)(D) in that case still bore some relevance to interpreting the term “representatives” in (a)(1).  Mr. Rothenberg again disagreed, stating that the discussion was now focused on what he believed to be the fundamental error of the district court — namely, treating all four characteristics of section (a)(2) as mandatory, because that would exclude otherwise able practitioners from representing taxpayers before Treasury simply because they lacked advocacy skills.  He also noted the position taken in the amicus brief of former IRS Commissioners that the “presenting a case” language could encompass preparing a tax return, but Judge Sentelle retorted that this would be an “awfully strange” use of the language.

Mr. Rothenberg then closed his argument by reiterating the government’s position that the district court erred in reading (a)(2) as limiting the language of (a)(1) and that (a)(1) itself did not foreclose Treasury from regulating tax return preparers.  Therefore, Chevron deference is owed to those regulations.

Counsel for the plaintiffs, Dan Alban, began his argument by maintaining that there was no statutory authorization for the regulation.  He described the statute as clearly focused on Treasury’s controversy and adjudicative functions, such as examination of returns and appeals before the agency, and not on what he described as “compliance” functions like filing a tax return.  He also pointed to the “presenting their cases” language in subsection (a)(2), stating that no “case” exists until there is a dispute over the taxpayer’s return.  Judge Williams asked about evidence that the scope of the original 1884 statute was limited to claims that were being resisted by the government — that is, controversies.  Mr. Alban replied that it was clear that the statute was addressing claims that the claimants chose to bring, rather than a mandatory function like filing a tax return.  In addition, he noted, the legislative history indicates that these were “contested” claims and that the representatives were standing in the shoes of the claimants.  Here, by contrast, tax return preparers are not “representatives” before the agency.  Judge Kavanaugh then asked who the preparers are representing.  Mr. Alban replied that they are not representatives of anyone; they are just performing a service in assisting preparation of the return, but the taxpayer himself has to sign it.  He noted in that connection that tax return preparers are not required to obtain a power of attorney, unlike taxpayer representatives in agency proceedings.

The court challenged Mr. Alban when he argued that the “level of policy decision” here warranted caution in allowing an agency, rather than Congress, to implement this new regulatory regime.  Judge Sentelle noted that counsel couldn’t get much “traction” with that argument when the D.C. Circuit frequently deals with “sweeping regulations” that create major changes in the regulatory landscape.  Judge Kavanaugh observed that, even if counsel was merely stating that the significance of the change ought to color the court’s approach to finding ambiguity, the suggestion was unworkable because it is hard for a court to decide what is “major.”

Finally, Judge Sentelle asked about the impact of the Supreme Court’s recent decision in City of Arlington holding that Chevron deference is owed to an agency’s determination of the scope of its jurisdiction.  Mr. Alban stated that the decision was not directly applicable, but in any case the Supreme Court had made clear in that case the importance of seriously applying the limitations on Chevron deference.  Here, because the statute was not ambiguous, Mr. Alban stated, the government’s position fails at Chevron Step 1, and therefore no deference is owed.  Putting aside the discussion of broader administrative law principles, it was not apparent that any of the judges on the panel disagreed with the plaintiffs on that basic point regarding section 330(a).

Mr. Rothenberg began his rebuttal with a general discussion of Chevron principles, stating that all the prior cases in which the D.C. Circuit had invalidated regulations at Chevron Step 1 were situations where the agency action was more clearly foreclosed by a specific Congressional determination found in the statutory text, but he was met with considerable resistance.  The judges observed that his list did not appear to be “exhaustive.”  In particular, Judge Sentelle suggested that this case was perhaps analogous to the American Bar Ass’n case, which he described as invalidating FTC regulations directed at the legal profession on the ground that Congress had not empowered the FTC to regulate that profession.  When Mr. Rothenberg answered in part that Chevron Step 1 sets a “low bar,” Judge Kavanaugh disagreed, stating that a court is to use all the tools of interpretation at Step 1 and that City of Arlington did not reflect a “low bar.”

Finally, Judge Kavanaugh asked Mr. Rothenberg to respond to Mr. Alban’s point that the IRS does not require tax return preparers to obtain a power of attorney.  He replied that the power of attorney is required for “agents,” and tax return preparers are not agents.  Mr. Rothenberg then repeated the point made in his opening remarks that the original 1884 statute covered “agents,” but other persons as well.  Judge Williams interjected that the government appeared to be placing too much weight on the statutory reference to “other persons,” because canons of statutory construction provide that the scope of broad language like that is limited by the specific terms that precede it — here, “agents” and “attorneys.”  Mr. Rothenberg noted that he disagreed, but his time expired before he could elaborate.

The case was heard on an expedited schedule, and therefore it is reasonable to expect that a decision will issue in the next couple of months.

Attached below is the plaintiffs’ response brief and the government’s reply brief, which were not previously posted on the blog.  The government’s opening brief and two amicus briefs in support of the government were previously posted here and here.

Loving – Plaintiffs’ Response Brief

Loving – Government Reply Brief

 

 

Next Page »