This news analysis was originally published in Tax Notes Today on January 15, 2008.
In December 2003 Sen. Barack Obama, D-Ill., was a little-known state lawmaker in Illinois, the Boston Red Sox had yet to shake off the Bambino's curse, and crude oil was selling for $28 a barrel. It was a different world.
That same month, the Senate Finance Committee launched an investigation into the IRS's advance pricing agreement program -- an investigation prompted by a noticeable increase in the number of APA requests from corporate taxpayers. The details surrounding the Senate's investigation were never revealed, but according to press reports, a draft report was complete as of June 2005. Excerpts leaked to the press revealed that there were problems with the APA program, including accusations that the program was poorly managed and that the IRS was giving away the store by agreeing to terms highly favorable to big business. (For prior coverage, see Doc 2005-13514 or 2005 TNT 119-1 2005 TNT 119-1: News Stories.)
Although neither the official draft report nor any subsequent report was ever released, Senate Finance Committee Chair Max Baucus, D-Mont., then the ranking minority member on the committee, publicly responded to the rumors, saying, "I encourage interested parties to wait and review the full report before drawing any conclusions from the APA program."
More than two years later, those same interested parties -- and the rest of the taxpaying public -- are still awaiting the results. No draft or final report has been issued, and no updates on the investigation have been announced. The Finance Committee, which Baucus now chairs, did not respond to Tax Analysts' repeated inquiries about the status of the report and whether the investigation is still pending.
In short, the Senate APA investigation has vanished into thin air.
Its unexplained disappearance raises several questions: How aggressively does the IRS negotiate APAs with corporate taxpayers? Do foreign governments stonewall the U.S. competent authority to gain concessions in bilateral APAs? Does the IRS quantify the dollar amount of controlled intercompany transactions covered by each APA? Do U.S. tax officials fear an outcry from corporate multinationals who have finally found a way around their transfer pricing headaches? And what does this episode say about the enforceability of separate company accounting and the arm's-length method?
We're no fans of conspiracy theories, but you can't help wondering: Is this a case of innocent bureaucratic fumbling, or did somebody kill the APA investigation?
The Evolution of APAs
By way of review, an APA is an agreement between a taxpayer and the IRS that establishes an appropriate transfer pricing method to apply to a transaction at issue. Unilateral APAs refer to agreements between the IRS and taxpayers, while bilateral APAs refer to agreements between the IRS, the taxpayer, and one or more foreign tax authorities.
APAs are voluntary and are entered into at the request of the taxpayer. Generally, they are prospective and cover the transaction at issue for a period of time, usually five years. In recent years, some APAs have included rollback provisions that allow taxpayers to apply them to past years.
In general, APAs are most frequently used by taxpayers in large transactions or in transactions in which the taxpayer anticipates controversy. In such cases, the taxpayer can negotiate with the IRS and lock in a pricing method through an APA to ensure that the transaction will not end up in an audit and result in a large adjustment.
In 1991 the IRS issued Rev. Proc. 91-22, establishing the procedures for granting APAs. In that same year, it granted the first APA to Apple Computer Inc. The revenue procedure does not address the motivations for APAs, but it was the first official guidance on APAs.
Since then, hundreds of APAs have been issued, and many more are in the works. As of October 5, 2007, the APA program had an inventory of 251 open APAs, of which 210 were bilateral and 41 were unilateral. In 2006 the IRS completed 82 APAs and 62 recommended negotiating positions; the estimates for 2007 are about the same. From the program's inception in 1991 through 2006, 692 APAs have been executed out of 1,037 applications filed, according to statistics taken from the IRS publication Announcement and Report Concerning Advance Pricing Agreements, issued on February 26, 2007. (For the most recent IRS quarterly report on the APA program, see Doc 2007-23055 or 2007 TNT 200-19 2007 TNT 200-19: Other IRS Documents.)
Everybody Loves APAs
Perhaps the most interesting aspect of APAs is that taxpayers and the IRS both seem to love them. In a world where those two sides often vigorously oppose each other, seldom does one find an area where rivals so clearly converge.
But they do. In interviews with Tax Analysts, several practitioners touted the virtues of APAs, complaining only that the process is time-consuming and costly but quickly asserting that the IRS is aware of those complaints and is making every effort to address them.
The argument for APAs is basically that transfer pricing involves heavy factual analysis, for which there is a lot of room for disagreement and the possibility of more than one correct result. Therefore, the ability to resolve those issues in a nonadversarial and prospective way is an effective way of administering the transfer pricing rules.
The APA program is also known for offering taxpayers a high degree of flexibility in resolving technical issues. That aspect of the program supports the cliché that transfer pricing is more art than science. One need look no further than the two public hearings the IRS devoted to APAs in February 2005 to get a sense of the importance that practitioners attach to the program's flexibility. The overarching theme of the hearings was taxpayers' pleas that the IRS not adopt a one-size-fits-all approach. (For the transcripts of those hearings, see Doc 2005-3085 or 2005 TNT 32-19 2005 TNT 32-19: Public Comments on Regulations; see also Doc 2005-4683 or 2005 TNT 45-22 2005 TNT 45-22: Testimony Other Than IRS and Treasury.)
That same flexibility is the source of one of the program's perceived drawbacks -- the increased likelihood of inconsistent treatment among similarly situated taxpayers, an issue that is likely aggravated by the fact that APAs are negotiated, and kept, in the dark.
But practitioners with years of APA experience do not necessarily agree with the view that APAs lead to inconsistent treatment. When the question was posed to Rocco Femia, a partner with Miller & Chevalier, he answered, "Inconsistent in comparison to what?"
Femia, who was associate international tax counsel with the Treasury Department from 2003 to 2004, said transfer pricing is all about the facts and that every taxpayer's facts are different. In that sense, there may be some perceived inconsistencies, he said. But he added that bigger inconsistencies can be found in field examinations, especially given that some transfer pricing issues are examined while others are not. "Compared to that, the APA program is a paradigm of consistency, as at least all the issues are reviewed," Femia said.
Joseph Andrus of PricewaterhouseCoopers agreed, saying that "results in the APA program are far less inconsistent than any other mechanisms that exist to resolve transfer pricing issues." Audits create far greater opportunities for either side to take unprincipled views of what the right outcome is, he said.
The IRS also loves APAs. As David Bowen, transfer pricing expert with Grant Thornton, explained, before the APA program, the only alternatives were appeals and litigation, and the IRS usually lost in litigation. "When viewed in that light," Bowen said, "the APA program is an attractive alternative for the IRS. That is, even a bad deal for the IRS beats getting blanked in litigation."
Despite the allure of APAs, they remain largely mysterious to the general public. Some groups have tried to make APAs publicly available, arguing that the agreements are equivalent to private letter rulings and other taxpayer determinations. Opponents of disclosure counter that APAs are substantively closer to IRS appeal settlements, which are confidential.
In 1996 the Bureau of National Affairs filed a Freedom of Information Act suit in U.S. district court requesting that APAs be made available to the public. (Tax Analysts, publisher of Tax Notes Today, supported the suit.) The FOIA suit eventually culminated in the IRS's conceding that APAs were "written determinations" under IRC section 6110 and agreeing that they would be made available for public inspection in redacted form.
The effort to make APAs public didn't succeed, however. Business groups mounted an impressive lobbying effort, and Congress quickly responded by amending the nondisclosure provisions of IRC section 6103 through legislation signed into law during the administration of President Bill Clinton. The Ticket to Work and Work Incentives Improvement Act of 1999 created section 6103(b)(2)(C), which classified APAs as confidential return information and therefore not available for public disclosure. The law also required that Treasury prepare annual reports on the status of APAs, and it was made retroactive to 1991, the year the APA program was launched.
The confidentiality protecting APAs means that few people know exactly what they contain. To date, only a single APA has ever been made public. In 2006, Tax Analysts obtained and published an APA related to litigation involving pharmaceutical giant GlaxoSmithKline. The APA was entered into in 1993 by the IRS and SmithKline Beecham, a predecessor of GlaxoSmithKline, and concerned the transfer pricing methods to be applied to intragroup sales of a popular antacid drug. (For the APA in the Glaxo litigation, see Doc 2006-19446 or 2006 TNT 180-78 2006 TNT 180-78: Other IRS Documents.)
From that APA and the guidance available on APAs, it is clear that APAs settle transfer pricing methods and provide the terms and adjustments to apply in determining the transfer pricing of the covered transactions. But the public doesn't know and can't get information about the specific terms granted or the basis on which they are granted. And the terms of two APAs cannot be compared. With such secrecy surrounding the negotiation of APAs, it is not clear whether similarly situated taxpayers are getting different results, whether different standards are being applied, or whether the IRS is negotiating a sweeter deal than what would have been available in the absence of an APA.
That's not to say that APAs should be made available for public scrutiny. There are legitimate arguments for keeping APAs confidential -- trade secrets and other information integral to a business's operation should not be forcibly revealed to competitors. However, given the large amounts negotiated in APAs, which often involve very large taxpayers, that APAs are secret adds merit to the argument that at the very least, the program should be subjected to government scrutiny and oversight.
As Bowen said, "Consistency of policy, practice, and procedure will always be the program's greatest challenge -- and that is why oversight is so necessary."
Speak No Evil
If everybody loves APAs and they work so well, why does the program need to be investigated? Furthermore, why would anyone be afraid of a report revealing the findings of such an investigation?
The answer to that is not clear, but Bowen, who was chief of APA Branch 3 in 2001-2002, said that the investigation was halted because "first, the report was informally lobbied into submission. Second, the Republican [Senate Finance Committee] which initiated the inquiry, went from majority to minority. After that, the report died for all practical purposes."
But why would anyone lobby against an investigation if the program is running so smoothly? Possibly because, as the leaks of the draft report indicated, the APA program is poorly managed and the IRS loses revenue by cutting deals for substantially less than would result from a transfer pricing adjustment. One reported recommendation in the draft report was to have the IRS quantify in dollars the amounts attributable to transactions covered by the APA. That would in turn provide a more defined sense of how much revenue the IRS is losing, if any, through the APA program, and possible evidence that the IRS is giving away the store.
However, it is difficult to prove that the IRS is losing revenue with the APA program. Andrus said that in his experience, the IRS is "very thorough and defends the interests of the government very well. They take their job seriously and they do a good job of defending the revenue." And as Bowen pointed out, in comparison to litigation in which the IRS seldom collected anything, collecting something may be better than nothing.
As for taxpayers, it's possible that they don't want a report that rocks the APA boat because the program allows them to cut deals with the government for a portion of what they are supposed to pay while avoiding the risks and costs of possible audits and litigations.
But is that really what's going on? At the end of the day, APAs are still costly and time consuming. Although audits or litigation may be more expensive than APAs, many taxpayers don't rely on APAs as their paramount transfer pricing planning tool. With many taxpayers relying on strong transfer pricing studies, APAs are often used only as a last resort when controversy is expected.
So while some taxpayers may be cutting sweet deals, many view the APA program as a way to resolve "important questions for which there is a lot of room for disagreement about what the right results are, and for which it's very clear that litigation is the wrong way to resolve them," Andrus said.
In the end, observers can only speculate on the basis of the leaked excerpts of the draft report. The more important question may not be why the investigation was stopped and by whom, but whether the program should be investigated to begin with.
"When one considers the dollar amounts that typically are involved in the average APA, it is more than reasonable to expect the agency charged with administering the program to voluntarily and willingly submit to some sort of effective and constructive oversight," Bowen responded.
Laying It All Out
It's impossible to say what specific aspects of the APA program the Senate investigation would have scrutinized, as the draft report has gone missing and no final report was ever prepared. Nevertheless, based on the letter that first got the ball rolling -- namely, a December 22, 2003, missive from Baucus and then-Finance Committee Chair Chuck Grassley, R-Iowa, to then-IRS Commissioner Mark Everson -- we have some idea of where the inquiry was headed. (For the letter, see Doc 2003-26954 or 2003 TNT 246-28 2003 TNT 246-28: Congressional Tax Correspondence.)
The committee leaders were clear about their motivation, reminding Everson in the opening paragraphs that "Congress has long been concerned about multinational corporations paying their fair share of tax." The letter asks the IRS to provide answers or information about 17 specific features of the APA program, and about enforcement under IRC section 482 generally. Those issues likely would have represented the guts of the Senate inquiry. In abbreviated form, the letter's discussion points are as follows:
1) For each company that received an APA during the last 10 years, which transfer pricing methods were proposed by each party and eventually used in the APA contract? And how much federal income tax, relating to the activities covered by the APA, was actually collected from the taxpayer?
2) For each company listed in question one, provide the cycle time for finalizing the agreement.
3) For each company listed in question one, indicate whether a bilateral APA was originally requested, and whether a bilateral APA was ultimately signed.
4) APAs often require taxpayers to provide an annual report to the IRS; provide all instances in which the annual reporter requirement was waived or modified.
5) For the last three years, identify all APA requests that were rejected by the IRS or withdrawn by the taxpayer.
6) Provide the APA program fees paid by each company listed in question one.
7) Identify all professionals currently involved with the APA and competent authority staffs and provide a summary of all disciplinary actions brought against these professionals; and explain why the APA program is under the jurisdiction of the Office of the Associate Chief Counsel (International).
8) Identify all professionals involved with the APA and competent authority staffs during the last 10 years and note which professionals have left employment with the IRS, and list all instances in which these former IRS professionals have contacted current APA or competent authority personnel regarding a proposed APA.
9) How much federal tax revenue is lost annually as a result of abusive or improper transfer pricing practices or income shifting?
10) How does the amount of U.S. taxes paid by foreign-controlled U.S. corporations compare with taxes paid to foreign governments by U.S. companies operating overseas?
11) How does the rate of return on assets of controlled foreign corporations compare with the rate of return on assets of domestic corporations?
12) How many CFCs paid no federal income tax for the last five years? How many domestic corporations paid no tax in these years?
13) Provide data and analysis of the IRS's "sustension rate" in transfer pricing cases.
14) Provide data and analysis of the IRS's success rate in the U.S. Tax Court and other courts involving transfer pricing disputes, including information on the total dollar value of tax adjustments proposed under section 482 that have been litigated and how much has been sustained by the courts.
15) Provide data and analysis of the total amount of tax underpayments by U.S. subsidiaries of foreign-owned corporations and how much of that underpayment involves transfer pricing issues.
16) Provide data and analysis of how many audit examinations the IRS has conducted since 1990 involving transfer pricing issues, how many of those examinations have been concluded, and all instances in which a taxpayer requested an APA after an examination was initiated.
17) Provide the updated tax return data on the CFCs, previously analyzed by the House Ways and Means Oversight and Government Reform Subcommittee during its 1990 review of U.S. transfer pricing issues, which earned billions of dollars in gross sales but reported no corporate profits and paid no taxes.
Regrettably, we have no idea how the IRS answered those questions -- if they answered them. But knowing the questions that Senate investigators were asking can bring observers much closer to understanding why the Senate investigation never came to fruition. It's easy to imagine that the penetrating nature of the questions directly contributed to the inquiry being scuttled. The sensitive topics range from preferred transfer pricing methods (no. 1), the APA program's revolving door (no. 8), and the estimated revenue drain from transfer pricing abuses (no. 9), to IRS difficulty in litigating transfer pricing disputes (no. 14).
The Bottom Line
The letter, the leaked excerpts of the draft report, and the missing final report all seem to point to the fact that someone killed the APA investigation. But in the interest of fairness, we must consider the alternative explanation -- that the investigation went away because the issues surrounding the APA program have been addressed since 2003.
There have, in fact, been changes introduced to the program in recent years. In May 2005 IRS Chief Counsel Donald Korb announced four reforms to be implemented in the APA program. Some reforms were designed to increase procedural efficiency (such as a requirement that parties to a bilateral APA sign a case schedule agreeing to complete an APA by the target date, with failure to do so requiring submission of a joint status report). Other initiatives, however, could be perceived as a step toward addressing some of the more sensitive issues surrounding the program. For example, one change introduced more lines of specialization, adding five new specialized groups to focus on the automotive sector, the pharmaceutical sector, the semiconductor sector, cost-sharing issues, and financial products. That, coupled with new hires, including a team leader and economist, indicates that the IRS is focusing on a higher degree of specialization, a change that could lead to more consistency in standards and taxpayer treatment and address the problem with the revolving door. (For coverage of Korb's announcement, see Doc 2005-9445 or 2005 TNT 86-3 2005 TNT 86-3: News Stories.)
But could that have been enough to stifle the load of very sensitive concerns one can infer from the leaked excerpts and the letter? Probably not. Those changes don't address the revenue concerns or the secrecy of the program. In fact, in the same speech, Korb reiterated that there were no plans to increase the transparency of the program.
Still, the reforms may be viewed as a step in the right direction and as enough to stop the investigation for the time being. There is no way to know for sure because nobody's talking, but at least it's a less conspiratorial explanation. The bottom line is that all we have is hearsay and an almost three-year-old promise from a politician to "wait and review the full report before drawing any conclusions from the APA program."
So, wait we shall.