The new base erosion and antiabuse tax (BEAT) in section 59A applies to payments by large U.S. corporate taxpayers to foreign related parties. There is an exception for certain services provided by foreign related parties.
There is no debate that this exception applies to services if the U.S. corporation paying its foreign related party for those services reimburses only the costs of services provided by the foreign related party. But there is debate about whether the exception from the BEAT applies to payments for services if those payments include a markup component. Specifically, does the BEAT apply to just the markup component of a marked-up service or to both the markup and cost component of the payment? For example, if a U.S. taxpayer pays $120 for services, reimbursing costs to the provider of $100 and generating $20 of profit to the provider, is $120 or $20 subject to the BEAT?
In a previous article, I took the position that $120 would be subject to the BEAT (“Can Marked-Up Services Skip the BEAT?” Tax Notes, Feb. 5, 2018, p. 705). In response, Manal Corwin, Ron Dabrowski, Michael Plowgian, Danielle Rolfes, and Thomas Wessel of KPMG LLP (hereinafter the KPMG authors) argued that my analysis was incorrect and that only $20 should be subject to the BEAT (“A Response to an Off-BEAT Analysis,” Tax Notes, Feb. 12, p. 933).
Plain Meaning
The relevant final statutory language (now code section 59A(d)(5)) describing the exception for services reads:
(5) EXCEPTION FOR CERTAIN AMOUNTS WITH RESPECT TO SERVICES Paragraph (1) [describing base erosion payments] shall not apply to any amount paid or accrued by a taxpayer for services if —
(A) such services are services which meet the requirements for eligibility for use of the services cost method under section 482 (determined without regard to the requirement that the services not contribute significantly to fundamental risks of business success or failure), and
(B) such amount constitutes the total services cost with no markup component.
The Senate statutory language did not include the last word “component.” The joint explanatory statement released with the final statutory language on December 15 has a different explanation for the Senate bill and for the final bill of the last phrase in the above statutory language. The explanation of proposed section 59A(d)(5)(B) in the Senate version reads: “and such amount constitutes the total services cost with no markup.” The corresponding description for the final law reads: “and only if the payments are made for services that have no markup component.”
In my article, my line of thinking was that the statutory text was ambiguous, and if that was the case, the legislative history performed the useful function of clarification. I would now like to make an additional observation about the statutory language. If the total payment for services can be bifurcated, it makes sense to me that if payment 1 is total cost ($100) and payment 2 (or payments 2a, 2b, etc., which add to $20) is profit, then the statutory language allows payment 1 to qualify for the exception and therefore not be a base erosion payment. On the other hand, if there is only one payment for the services in question and that payment contains a markup component, it clearly does not qualify for the exception.
So we now must focus our attention on whether the language allows for bifurcation of payments for services. The key phrase is “any amount . . . paid for services.” In everyday parlance when we refer to what we paid for an item, we are referring to the total payment we make. Simply put, we are talking about the price. When we say we paid $2 for a cup of coffee, we are referring to the total cost to us as the purchaser. If you want to get absurdly precise, we could say about the $2 coffee that we paid the 50 cents it cost to make the coffee as part of the cost. But to say we paid 50 cents for a cup of coffee because 50 cents is a component of the total payment for the coffee would be misleading. In fact, it would be so misleading that it would have the same effect as being untruthful. In any unrelated-party context, we would not normally have access to the information on pricing and profit splits. All a buyer needs to know is the amount of money they must pay. How this amount is calculated is irrelevant.
It is not just in everyday language where this interpretation of the language prevails. In the context of transfer pricing of services in the regulations when reference is made to “amounts paid,” the plain meaning is for the total amount paid in a related-party transaction and not to any component. The very first sentence of reg. section 1.482(9) states: “The arm’s length amount charged in a controlled services transaction must be determined under one of the methods in this section” (emphasis added). Here, this means the price of the services (charged by the provider and paid or accrued by the purchaser). As the text of the regulations describes, this amount can be total costs under the services cost method (SCM), a market price under the comparable uncontrolled services method, or a cost-plus markup under other methods. There is no connotation that the amount can or will be divided into different payments or accruals. Why should there be? Divisions (to the extent they occur at all) of the total price paid by a payer to a payee are economically meaningless. It seems unlikely that drafters would intentionally design a statute in which so much depended on a form that has no meaning anywhere else in the code or business practice.
If we interpret the BEAT services exception as allowing bifurcation, and payments and accruals for the cost component are further divided (into payments 1a, 1b, etc., for example, because of some arbitrary quirk in a taxpayer’s accounting), those payments would not constitute total cost and therefore would not qualify for the exception. According to the regulations, “Total services costs means all costs of rendering those services for which total services costs are being determined.”
If we interpret the language as allowing bifurcation, note that taxpayers who keep payments and accruals integrated receive a harsh result relative to those that bifurcate. That such different results would occur because of something as trivial as the forms of payment (in this case, because it is a transaction between related parties, an accounting entry) that are economically equivalent is extremely poor policy. It would be a classic case of a trap for the unwary. Further, it is hard to see what economic, accounting, or business purpose is served for the purchaser, especially if the purchaser is operating under the arm’s-length standard, to engage in the peculiar practice of splitting payments between cost and profit of the seller.
Also note that under the bifurcation interpretation, we need a new definition of “amounts paid or accrued” that is different from the common language usage of total cost borne by the payer or under the language in the regulations. Under what circumstances is one amount paid or accrued distinguished from another? What is the implication for such a novel interpretation of “amount paid” for other areas of the tax law?
The Parenthetical
All references to the BEAT in the early and final statutory language, and in official explanations of those statutory drafts and of the statutory language of the final law, state that the definitions of services that are exempt from the BEAT must meet certain requirements of the SCM in the regulations. And those certain requirements exclude what is known as the “business judgment rule.” By removing this requirement, the new law expands the scope of services to which the BEAT applies to include services that contribute significantly to fundamental risks of business success or failure. It is certainly reasonable to generally expect these services to be marked-up services, especially — as the KPMG authors point out — because the preamble to the 2006 proposed regulations introducing the SCM stated in reference to the business judgment requirement that the “condition effectively is reserved to allow the IRS to reject any attempt to claim that a core competency of the taxpayer’s business qualifies as a mere back office service.” The KPMG authors make the logical argument that that the dropping of the business judgment rule is meaningless if, as explained in the statement accompanying the conference report, the exception is meant only to apply to services with a transfer price with no markup component because the business judgment parenthetical only applies to services with a transfer price that includes a markup component. To repeat the idea in somewhat simpler terms, if the general rule is that a category must only include apples, why include a statement that expands the definition to include more types of oranges?
At this juncture we provide an overview of the SCM. In the 2009 regulations, if four requirements are met, no markup is required for amounts paid for and received by related parties as payment for provision of related-party services. Those requirements are:
(1a) the service is a “specified service” on a white list of over 100 services (Rev. Proc. 2007-13, 2007-1 C.B. 295 ), or (1b) it is a “low margin covered service” for which there is a controlled services transaction that has a median comparable markup on total services costs that does not exceed 7 percent (collectively (1a) and (1b) define covered services);
(2) the service is not an excluded activity from a blacklist of activities (for example, manufacturing, production, and extraction);
(3) the service must pass the business judgment test (which maintains that the taxpayer must reasonably conclude, in its business judgment, that the service does not contribute significantly to key competitive advantages, core capabilities, or fundamental risks of success or failure of a trade or business of either the provider or recipient of the services); and
(4) adequate books and records are maintained.
Here’s one possible plain-English restatement. Services can be on the white list, the blacklist, or neither. From this starting point we must identify “covered services” because only they potentially qualify for the SCM. Covered services are all services on the white list plus all services on neither list if their markup is less than 7 percent. To qualify for the SCM, there are two further requirements: The covered services must pass the business judgment test, and the taxpayer wishing to apply the SCM must keep adequate books and records as described in the regulations. The table below is an attempt to illustrate all this.
Under reg. section 1.482(9), the SCM may or may not be the best method used in determining the amount paid for services in a related-party transaction. And under the new statute, for purposes of determining a base erosion payment, a taxpayer seeking to qualify for the exception for certain services need not be using the SCM. As the KPMG authors pointed out, the purpose of the reference to SCM regulations is to describe the services potentially qualified for the BEAT services exception, not to describe the amount paid for the services.
The KPMG authors argued that the extra services made eligible for the exception because of the exclusion of the business judgment requirement must have a markup. There is nothing (that I can see) in the text of reg. section 1.482(9)(b) that indicates this. What the regulations do say is that if services do not qualify for the SCM because (for example) they fail the business judgment rule, the taxpayer must determine the best method from among the other methods described in reg. sections 1.482(9)(c) through 1.482(9)(h).
The first of those methods is the comparable uncontrolled services price method. Under this method, if a taxpayer is also providing to an unrelated party services sufficiently comparable to those related-party services in question, the price charged (perhaps with adjustments) to an unrelated party should be used for the transfer price. Under this method, there is no guarantee of a profit margin simply because not all business with unrelated parties is profitable. Thus, the expansion of services eligible for the exception in the definition of base erosion payment is not meaningless. Some transactions with no markup determined under the comparable uncontrolled services price method (and perhaps other methods) are included in the exception by the removal of the business judgment requirement. Total amounts paid for a service that does contribute significantly to fundamental risks of business success or failure may not have a profit margin.
“Specified” (white list) | Neither “Specified” Nor “Excluded” | “Excluded” (blacklist) | ||||||
|---|---|---|---|---|---|---|---|---|
Examples of Markup Percentage: | Examples of Markup Percentage: | Examples of Markup Percentage: | ||||||
0% | 5% | 40% | 0% | 5% | 40% | 0% | 5% | 40% |
[1a] “Specified” service (technically can be zero, low-, or high-margin) (“generally do not involve a significant arm’s-length markup on total services costs”) — OK | [1b] “Low-margin” (less than 7%) — OK | “High-margin” (greater than 7%) — NO | [2] “Excluded” service — NO | |||||
[1] “Covered” service — OK |
| |||||||
[3] Business judgment rule (applies only to covered services): Not services that contribute significantly to fundamental risks of business success or failure — OK |
| |||||||
[4] Have adequate books and records — OK | ||||||||
Even if the parenthetical statement dropping the business judgment rule was determined to be meaningless, that does not necessarily require readers to conclude that the statute’s interpretation as described in the official explanation is wrong. The inclusion of the parenthetical could be interpreted as violating one of the canons of statutory construction that text should not be ignored as meaningless; in legal lingo, according to this canon, the text in parentheses cannot be dismissed as “mere surplusage.” However, statutory interpretation is a haphazard process — more art than science. There are very few strict rules when it comes to statutory construction in practice, and almost all approaches are subject to controversy among legal scholars. Meaninglessness (if that is the case) is not an error of logic. It is just language without any relevance. It happens a lot in the creation of software, where it is just considered inefficient programming and poor style. It happens a lot in writing for publications, and skillful editors remove it. But if redundancy is not removed, it does not change the intent of the authors.
Colloquy
I must also disagree with the KPMG authors on the significance of the colloquy between Sen. Rob Portman, R-Ohio, and Senate Finance Committee Chair Orrin G. Hatch, R-Utah. The degree of significance of colloquies in interpretation of tax law is highly controversial.
Legislative debate can be revealing to legislative intent, but prepared statements like the Portman-Hatch colloquy (which was not even delivered orally on the Senate floor but inserted into the record) are rarely considered valuable in interpreting legislative intent. Judge Abner J. Mikva, a former member of Congress, did not believe much weight should be given to a colloquy that he colorfully characterized as “a pas de deux where two members get up and read a congressional version of a psychodrama from a prepared script.” (See “Reading and Writing Statutes,” Texas Law Review, 1986, p. 187.) Compare this with the comments on legislative history by Frank B. Cross, who considers a conference committee report the most authoritative single source. He writes: “Of the hierarchy of legislative history sources, the content of the report of the conference committee stands at the apex” (The Theory and Practice of Statutory Interpretation, 2008).
Regardless of one’s general view of colloquies, it seems in this case that it is clear that the explanatory language (which would disqualify the $120 in our example) must prevail over the colloquy (which would only disqualify $20) for at least two reasons. First, the joint explanatory language was accompanied by a conference report approved by the entire conference committee of both houses of Congress, while the colloquy only involved two members of the taxwriting committee of one house of Congress without any formal indication of approval by any other members. Second, the joint explanatory statement came two weeks after the colloquy, and if the conferees in their explanation wanted to pick up on the point of the colloquy, they probably would have amplified it in some way instead of contradicting it.
Post-Enactment Staff Opinions
Lowest of the low in any court’s interpretation of statutory language is the opinion of staff expressed in the aftermath of the legislative process (even though they in fact are probably the only folks who know the true intention of the draft). This is underlined by the fact that all public utterances by staff are prefaced by the statement that they are only providing their own personal views and not those of their bosses. Nevertheless, though the courts may disregard staff opinions, Treasury seems willing to consider them, and both chairs of the taxwriting committees have publicly expressed their intent for Congress to play a significant role in Treasury’s interpretation of statutes as the department rolls out prodigious amounts of non-trivial guidance for the new legislation.
Tax Analysts publications have recorded the opinions of four current staff members who were involved in drafting the BEAT. In a panel discussion at the District of Columbia Bar Taxation Community on January 25, Loren Ponds, majority tax counsel for the House Ways and Means Committee, stated in response to a question: “As written in the statute, the entire amount . . . . So let’s say your cost is 100 and your markup is 7, the whole 107 is treated as a base erosion payment.”
On the same panel was Jennifer Acuna, majority senior tax counsel and policy adviser for the Senate Finance Committee. She also pointed to the conference report’s language to support the contention that the entire payment would be treated as a base erosion payment ( Tax Notes Int’l, Jan. 29, 2018, p. 418).
On February 15 at a Tax Council Policy Institute conference in Washington, there was a panel that included congressional tax staff. Among the members were Mark Prater, deputy staff director and chief tax counsel for the Finance Committee, and Barbara Angus, majority chief tax counsel for the Ways and Means Committee ( Tax Notes, Feb. 19, 2018, p. 1101).
It seems that in Prater’s view, regarding a payment for services that includes a markup, the Senate floor colloquy should be read in a manner that is consistent with the conference report language that succeeded it. He said he does not believe the colloquy between Portman and Hatch is by itself determinative of whether the BEAT applies to services that include a markup. “I advise people that it’s just a colloquy that discusses the issue, but if you look in the conference report language, it appears to be pretty clear that, in general, the exception discusses the services with no markup,” Prater said.
Angus was less direct and appeared to cast doubt on Prater’s view. “There is the statutory language, and there’s the legislative history,” she said. “But the colloquy is important. These are issues that the Treasury Department will look at now. . . . I would think this is a good example of places where we encourage people to come in and provide feedback.” As noted above, it is not clear to me why the December 1 colloquy is important when it is clearly overridden by the December 15 joint explanatory statement. And it is not clear why the provision of feedback from the private sector will help resolve the issue.
Senate Intent, Conferees’ Intent
It may assist readers’ understanding to frame the debate on this topic chronologically. There are several possible scenarios, not only regarding the final legislative intent but also possible changes in legislative intent over time. Table 2 lays out some of those possibilities. If legislative intent did in fact change over time, this would go a long way toward explaining the misunderstanding by some who perhaps do not take into account possible changes in intent.
The only way for scenario (1) to be a plausible interpretation of congressional intent is to ignore or totally discount the plain language of the explanatory statement accompanying the conference report that (to remind readers) states that the exception applies “only if the payments are made for services that have no markup component.” Scenarios (1) and (2) are consistent with the belief that the Portman-Hatch colloquy seals the deal. Under scenario (3), the Portman-Hatch colloquy is not a mere clarification but attempts to recharacterize the intent of the Senate statutory language, but the intent at the time of Senate passage is not sustained in conference. (In this case, the colloquy would be a way of providing taxpayers with last-minute relief without paying for that relief in the official Senate revenue estimates that were critical at the time.) Under scenario (6), the Portman-Hatch colloquy again is not a mere clarification but an attempt to change the intent of the Senate statutory language, and that clarification of intent is indeed consistent with the law enacted.
This article offers the view that scenarios (4) and (5) are most plausible. Under (4), the colloquy is a clarification of intent. Under (5), the colloquy is a clarification of an issue not previously considered. If, as is reasonable, we take the view as a general matter of statutory construction that the statutory text does not change except to signal a change in meaning or to clarify, the addition of the word “component” to the statutory text of the final conference version (and not in the prior Senate versions) can be interpreted as a signal of change or clarification of intent. Unfortunately, the addition of the word “component” does not tell us in and of itself what that signal or clarification may be. But the simultaneously published explanation of the Senate version and the conference agreement does.
Scenario | Intent of SCM Exception | ||
|---|---|---|---|
Marked-Up Services in Exception, Therefore May Not Be Subject to BEAT |
No Explicit Intent | Marked-Up Services Not in Exception, Therefore Subject to BEAT | |
(1) | Senate/Conference |
|
|
(2) | Conference | Senate |
|
(3) | Conference |
| Senate |
(4) | Senate |
| Conference |
(5) |
| Senate | Conference |
(6) |
|
| Senate/Conference |
Conclusion
Should taxpayers and Treasury interpret the new law in accordance with the official explanation? Or should one possible explanation of that language that is contradictory to the official explanation prevail? I would venture to say, despite some keen observations by the KPMG authors, that the preponderance of evidence from the written record is in favor of the explanation in the joint explanatory statement that is clearly saying that if a payment is made for a service by a U.S. corporation to a foreign related party, the whole of the cost is included in the BEAT unless the whole, entire, and complete price paid for the service reflects only the cost of the provider of the service. That’s what will matter if this issue goes to court.
For now we must be concerned with Treasury’s interpretation of the statute. (And after all, it is hard to see how the case will ever get to court if the IRS does not challenge taxpayers.) Input from congressional staff on intent will matter as well as the legislative history. On February 2 Dana L. Trier, Treasury deputy assistant secretary for tax policy, explained that implementing tax reform requires “a considerable amount of discussion . . . with our counterparts on the Hill” ( Tax Notes, Feb. 12, 2018, p. 960). And on February 14, in an opening statement at a Senate Finance Committee hearing, Hatch staked a position that one could hardly describe as deference to the executive branch: “I’m going to keep working to ensure that everyone recognizes and respects Congress’s role in this process and the fact that the best place to get an explanation of Congress’s intent is Congress itself. Where things are potentially unclear in the law, Congress should be the one to determine and explain what was intended.”
Based on public statements so far, the evidence strongly favors the view that prices for services that have a markup component are not included in the exception and therefore are base erosion payments subject to the BEAT.
Andrew Velarde and Alexander Lewis contributed to this article.