February 1998
Transfer Pricing and Confidentiality
Major Case on Work Product Privilege
Transfer Pricing and Confidentiality
One problem that has been lurking in the transfer pricing arena is the potential use by tax authorities of secret third party data to support transfer pricing adjustments. In the U.S., Section 6103 operates as a brake on such use because that information cannot be shared with the taxpayer under audit. For a more complete discussion of the problems encountered in the U.S., See Townsend, Section 6103 and the Use of Third Party Tax Return Information in Tax Litigation, 46 Tax Lawyer 923 (1993) (which covers the range of problems arising from the use of such information in transfer pricing disputes).
The U.S. competent authority has expressed some concern about negotiations with treaty counterparts where the treaty partner may be using third party information to support its position in the negotiations. Obviously, the competent authority process works only to the extent that each negotiating side understands the basis for the other’s position. This would necessarily mean that the U.S. must have access to the treaty partner’s otherwise secret information. Furthermore, since a U.S. taxpayer’s tax liability is involved, it must be expected that at some point the taxpayer might also access this information. If there are restrictions on this information ultimately getting to the taxpayer, there will be problems in the negotiating process. The U.S. competent authority has expressed grave reservations about the use of such information.
BNA’s Daily Tax Report has reported that the Mexican authorities are moving to allow the use of such secret information. See Official Says Mexico Can Control Access to Confidential Third Party Data, 1998 DTR 27 d10 (2/10/98). The background is that Mexico has less access to public source financial and related information from which to develop transfer pricing audits. Accordingly, Mexico perceives that effective transfer pricing compliance may require access to third party information. The solution Mexico developed is to allow two representatives of the taxpayer under audit to access financial information from unrelated third parties. The right to such access is accompanied by stiff jail penalties if the representatives misuse the information.
Taxpayers’ concerns are, of course, that it will be difficult to determine whether there is misuse of such information, either by the Mexican tax authority or by competitors. The Mexico authorities believe that the problem is not as large as it appears, because the information relates to earlier years and therefore is not as sensitive.
Further, the Mexico authorities indicate that, in a number of cases, the information can be derived by looking to the non-Mexican side of the transaction, particularly to the U.S. related party. The Mexico authorities state that 60% of the transactions are with U.S. entities where information is more readily available. If that information indicates an appropriate transfer price from the U.S. side of the transaction, it will perforce indicate an appropriate transfer price for the Mexico side of the transaction. (Perhaps the Mexican authorities will directly or indirectly push the audit work to the U.S. side of the border.)
Finally, the Mexican authority indicated that, in other countries (other than U.S.) facing similar limits on publicly available data, the tax authorities are considering similar approaches to developing proper transfer pricing information.
Townsend and Jones Comment: U.S. taxpayers, as well as taxpayers generally, conducting business in Mexico and other countries that use such third party information need to be concerned. Where that use is secret, of course, the taxpayer under audit may have a difficult time overcoming proposed adjustments. Where the taxpayer has some limited access, presumably the taxpayer will at least know what it is shooting at, but then the comparable taxpayers may have been cherry-picked by the tax authority to support its view of the universe. (It happens in the U.S. and it will certainly happen abroad.) The system makes sense only if the taxpayer is allowed some opportunity to review and use secret information of potential comparables of its choosing. And, in either event, the dissemination of secret information, even if somewhat dated, is bound to create competitive problems or at least the suspicions of abuse of the information. With heavy penalties such as those suggested by the Mexican authorities, we are not sure that we would want to be the taxpayer’s representatives having access to the information. On the other hand, taxpayers who are risk takers may want the opportunity to review their competitors' secret information in the belief that they can hide their competitive use of such information. It is a big problem that bears watching and caution as dealing with Mexican and similarly minded tax authorities..
Major Development in Work Product Privilege - 2d Circuit Expands Privilege
In US v. Adlman, ___ F.3d ___ (2d Cir. 2/13/98), the Second Circuit adopted a broad interpretation for the work product privilege. Our readers may recall this case which had been previously appealed to the Second Circuit. The background facts were: the tax director (an attorney) of a large corporation sought tax advice from the corporation's regular accounting firm (Arthur Andersen) as to a proposed transaction. The transaction was designed to generate a very large loss and refund. The memorandum assessed the potential litigation outcomes in the event of the virtually certain IRS challenge of the transaction if it were undertaken. Apparently the transaction was sufficiently aggressive and the likelihood of intensive IRS review of the transaction so great because of the size of the expected refund, that an IRS challenge was viewed as virtually certain.
In the earlier appeal, the issue was whether IRS summons issued to Adlman to produce this document could be enforced over his assertion, on behalf of the corporation, of the attorney-client privilege and the work product privilege. The Second Circuit then held that the attorney-client privilege did not apply because the corporation had not done the predicates necessary under Kovel and its progeny to qualify the accountant's work as legal advice. The Second Circuit then also held, however, that the district court had evaluated the work product privilege under the wrong standard. The Second Circuit remanded the case for consideration of the applicability of the work product doctrine.
On remand, the district court determined that the work product privilege did not apply, apparently applying test that the memorandum was prepared to evaluate a contemplated business transaction rather than principally for a litigation purpose. On the second appeal, the decision now being discussed, the Second Circuit held that the work product doctrine was broader than the district court had applied it. The Second Circuit held that the work product doctrine can apply if "in light of the nature of the document and the factual situation in the particular case, the document can fairly be said to have been prepared or obtained because of the prospect of litigation." (Citing the leading treatise of Wright, Miller & Marcus, and referring to this test as the Wright & Miller test.) Thus, even though a document is prepared to inform a business decision (such as the acquisition involved), it still would qualify for the work product privilege if the analysis focused upon the prospects of and litigation outcome an IRS challenge (or, presumably, other types of legal challenges),
In so holding, the Court aligned itself with several courts that had reached a similar interpretation of the work product privilege and rejected the holdings of other courts, including principally our circuit (the Fifth Circuit), that had adopted a more restricted interpretation of the work-product doctrine. In United States v. El Paso Co., 682 F.2d 530 (5th Cir. 1982), cert. denied, 466 U.S. 944, 104 S. Ct. 1927 (1984), the Fifth Circuit had rejected the application of the work-product privilege to an analysis of "prospective liabilities that might result from litigation with the IRS over [the taxpayer's] tax returns." The Court reasoned that the analysis was prepared for financial statement purposes and not in anticipation of litigation. The Second Circuit rejected the Fifth Circuit's interpretation. (Note that, given this conflict, the case may present the United States the opportunity to petition the Supreme Court for certiorari so as to resolve this very important conflict.)
In its holding, the Court cautioned that the work-product privilege is not absolute and can be avoided if the opposing party makes a showing of need for the document or information. Thus, it is critically important, as the first decision in Adlman taught us, that the taxpayer qualify the information or the document for the attorney-client privilege, something that could have been done easily in Adlman with attention to just a few details (specifically, the traditional Kovel agreement that a lawyer such as Adlman could have entered with the accountants). The attorney-client privilege is absolute and cannot be pierced. As noted, the work-product doctrine is qualified and can be pierced by a showing of need. In Adlman, the Second Circuit remanded for the district court to apply the proper test to determine whether the document qualified for the work-product privilege. In doing so, the Court expressly rejected the IRS's argument that remand was unnecessary because the IRS had made the required showing of need that could override the privilege. The Court held that the document in question contained "mental impressions, conclusions, opinions or legal theories of an attorney or other representative" (citing Rule 26(b)(3)), the quintessential type of information toward which the work-product doctrine was directed, requiring at a minimum a "highly persuasive" showing of need. The Court found that the IRS simply had not made the showing. (To which we say, what possible relevance to the tax characterization of the transaction would it make what the accountants thought and recommended about the transaction; on the other hand, it might be relevant if the IRS were considering penalties but penalties would be highly unlikely where, as in Adlman, the tax benefits of the transaction were claimed on refund claims certain to be reviewed by the IRS and Congress (under the Joint Committee approval requirement).)
T&J Notes: As noted, the Fifth Circuit has not yet expanded the work-product privilege in this manner. Thus, in the Fifth Circuit, it is particularly important that taxpayers qualify for the attorney-client privilege through the simple steps noted above.