How much is enough? Lessons on transfer pricing documentation from the recent IRS report.

• Author: Durst, Michael C.

Introduction

In May 2002, the IRS released its report to Congress on the effectiveness of the "contemporaneous documentation" that the Code and Regulations have, since 1994, required multinational companies to prepare in order to avoid exposure to penalties in the event of transfer pricing adjustments under section 482. (1) The report is based on a study by IRS examiners of documentation that had been submitted by 25 different taxpayers, as well as a survey of taxpayers conducted on behalf of the IRS by an independent consultant, Schulman, Ronca & Bucuvalas, Inc. (SRBI). Although it can be questioned whether the survey was truly representative, the report nevertheless provides useful information for tax executives seeking a cost-effective approach to documentation. In particular, the report supports the following practical conclusions:

(i) Contemporaneous documentation has become part of the fabric of tax practice, and IRS examiners expect documentation to be available at the start of large case examinations. Companies that do not maintain documentation with respect to their significant intercompany transactions are a conspicuous minority and face a much-heightened risk of penalties.

(ii) The IRS report indicates that documentation is removing some difficulty from the early stages of examinations by reducing the number of IDRs that are issued. The report does not suggest, however, that documentation is reliably reducing the contentiousness of examinations once the IRS identities a transfer pricing issue. Even the best documentation is not likely to eliminate controversy if the IRS otherwise is inclined to inquire into pricing.

(iii) Documentation is proving costly to multinational companies. Indeed, companies may be spending more than necessary on the documentation that they are producing, and could consider reducing costs by avoiding unnecessarily extensive analysis, and by doing more of the necessary work in-house.

Background of the Report

The recent report seems to have originated in the longstanding concern of Senator Byron Dorgan that the arm's-length standard cannot serve as the basis of an effective transfer pricing system. At Senator Dorgan's request, (2) the Senate Finance Committee included, in its version of what later became the fiscal year 2000 Treasury-IRS appropriations legislation, a requirement that the IRS conduct a study of the extent to which taxpayers are preparing documentation, and the quality of the documentation that is being submitted. (3) The conference committee later deleted this requirement, but the IRS decided nevertheless that the report should be provided. (4)

Methodology

SRBI conducted its taxpayer survey primarily during 2001, by initial telephone contact followed by an Internet-based questionnaire. (5) The IRS took pains to reassure companies that no data identifiable with particular taxpayers would be given to the IRS. (6) Despite these precautions, however, when the survey firm began contacting taxpayers with its questionnaire, the tax executives of many companies faced conflicting considerations concerning whether to participate. Notwithstanding the precautions taken by the IRS to ensure confidentiality, some potential participants thought it undesirable to record in an unprivileged setting observations concerning their companies' relative degree of compliance with documentation rules. (7)

On the other hand, it seems likely that some tax executives did not want, by declining to participate in the survey, to send Congress an implicit message that the arm's-length standard as currently implemented is not working. Given the history of the survey, it was natural to view the effort as a "report card" on the operation of current transfer pricing rules. As it turned out, of those companies that the survey firm identified as potentially eligible for participation in the survey, about 70 percent actually participated. Of the 30 percent that opted out, about half explicitly declined to take part, and the other half simply did not send in completed forms. (8)

This rate of nonparticipation does not in itself cast doubt on the validity of the sampling, but it seems likely that the self-identification of the 70 percent who did choose to participate was not entirely random. Those who did participate probably represent disproportionately companies that both (i) felt that their own documentation was in particularly good order, and (ii) desired to demonstrate for the benefit of Congress the effectiveness of current transfer pricing rules. (9) Thus, this possible skewing suggests that the report may present a somewhat rosier picture of documentation practice today than would be revealed by a more rigorously representative sampling.

Similarly, the IRS's own study of 25 taxpayers' documentation must unavoidably be seen more as an approximate survey than as a statistically rigorous exercise. The process used to select the 25 taxpayers was not designed (and probably could not practically be designed) to select a fully representative sampling. (10) Moreover, there is no quantitatively precise way to judge the quality of documentation. Examiners therefore were asked to grade documentation into four categories--excellent, good, moderate, and poor--and there can no assurance that the standards applied by the different examiners that conducted the survey were consistent. Nevertheless, given the limitations on the precision that any survey of the quality of documentation could achieve, the IRS effort was conducted carefully and provides a useful body of information.

Implications of the Surveys

Documentation Is Here to Stay

The survey strongly suggests that a large majority of multinational companies in fact prepare documentation for those transaction flows that might reasonably become the subject of disputes with the IRS involving the possible assertion of penalties. (11) There can be little doubt that the documentation rules are accomplishing at least one of the goals intended by Congress in enacting them--namely, to encourage companies to give careful consideration to their transfer pricing policies, and to whether those policies can be explained and defended successfully on examination. Certainly, those large multinational companies that have not yet established an organized approach to identifying situations in which documentation appears necessary, and that are not compiling the documentation, are putting themselves at risk.

High Costs; Limited Utility of Documentation in Examinations

The survey also suggests, however, that companies are spending large sums on documentation. One is left wondering whether Congress intended the resource demands of the documentation system to be quite so high, particularly in relation to the relatively modest utility of the documentation. The survey suggests that, of those companies that had prepared documentation and responded to the survey question about staff-resource commitments, 80 percent devote at least one full-time equivalent (FTE) to "documentation and other transfer pricing issues. (12) Another and somewhat overlapping question asked how much taxpayers had spent on both internal and external resources in producing the documentation (excluding, for some reason, the cost of producing annual updates). Of those responding to the question, 35.6 percent reported spending $100,000 or more, with the percentage increasing to 55.6 percent for respondents with gross receipts of more than $1 billion. (13) Of the respondents with gross receipts greater than $1 billion, 26.1 percent reported spending from $200,001 to $500,000, 6.8 percent from $500,001 to $1 million, and 2.3 percent more than $1 million. (14)

The survey does not provide a confident basis for estimating the total amount spent on U.S. documentation, but can be read to suggest a total cost of at least $250 million and probably much more (with about $500 million being our admittedly subjective best guess). (15) Whether the amount spent is excessive is, of course, a matter of subjective judgment. It should be remembered, however, that the U.S. corporate income tax nets about $240 billion in revenue annually. Although compliance costs represent real expenditures of resources, tax collections do not in themselves represent a net increase in economic resources, but rather a transfer of resources from private to public use. (16) It seems anomalous for society to expend, in real resource costs, something between an eighth and a quarter of one percent of these revenues transferred from private to...