United States v. Textron Inc.

Opinion

United States Court of Appeals For the First Circuit

No. 07-2631 UNITED STATES,

Petitioner, Appellant,

v.

TEXTRON INC. AND SUBSIDIARIES, Respondent, Appellee.

APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF RHODE ISLAND

[Hon. Ernest C. Torres, U.S. District Judge]

Before

Torruella, Boudin, Circuit Judges, and Schwarzer,* District Judge.

Kevin J. O'Connor, with whom Nathan J. Hochman, Assistant Attorney General, Richard T. Morrison, Deputy Assistant Attorney General, Gilbert S. Rothenberg, David I. Pincus, Robert W. Metzler, Judith A. Hagley, Attorneys, Tax Division, Department of Justice, and Robert Clark Corrente, United States Attorney, was on brief for appellant.

John A. Tarantino, with whom Patricia K. Rocha, Adler Pollock & Sheehan P.C., Arthur L. Bailey, J. Walker Johnson, and Steptoe & Johnson LLP, was on brief for appellee.

David M. Brodsky, Robert J. Malionek, Adam J. Goldberg, Latham & Watkins LLP, Robin S. Conrad, Amar D. Sarwal, Susan Hackett, Senior Vice President and General Counsel, Attorneys for the Chamber of Commerce of the United States of America and Association of Corporate Counsel, as amicus curiae in support of Textron Inc.

*

Of the Northern District of California, sitting by designation.

Kevin L. Kenworthy, Alan I. Horowitz, and Miller & Chevalier Chartered, Attorneys for Financial Executives International, as amicus curiae in support of Textron Inc.

January 21, 2009

3

TORRUELLA, Circuit Judge. The question presented by this appeal is whether the work-product doctrine protects documents prepared by Textron Inc. for the purpose of calculating tax reserve liability from production to the IRS pursuant to an investigative subpoena. Like other companies, Textron prepares "tax accrual workpapers" which, generally speaking, list the questionable positions Textron took on its tax returns, estimate the likelihood that those positions will not withstand scrutiny, and calculate the amount of additional tax liability that would result from revision of those positions. Textron prepares these estimates so that it can maintain an adequate reserve fund, properly report its assets and liabilities, and obtain independent certification of its financial statements. As part of the auditing process, Textron showed these tax accrual workpapers to Ernst & Young ("E&Y") an independent auditor.

This case arose when the Internal Revenue Service ("IRS"), after noticing potential tax shelter transactions, issued an administrative summons to Textron pursuant to I.R.C. § 7602 seeking tax accrual workpapers for Textron's 2001 tax returns. Textron refused to comply and asserted a number of defenses. The IRS sued to enforce the subpoena. After an evidentiary hearing, the district court ruled for Textron on its work-product protection claim, but rejected its other defenses. The district court also found that Textron's disclosure to E&Y did not constitute waiver.

4

The IRS appeals. After careful review,1 we affirm in part, vacate in part, and remand.

I. Background

The IRS subpoena sought "Tax Accrual Workpapers," defined as:

[A]11 accrual and other financial workpapers or documents created or assembled by the Taxpayer, an accountant for the Taxpayer, or the Taxpayer's independent auditor relating to any tax reserve for current, deferred, and potential or contingent tax liabilities, however classified or reported on audited financial statements, and to any footnotes disclosing reserves or contingent liabilities on audited financial statements. They include, but are not limited to, any and all analyses, computations, opinions, notes, summaries, discussions, and other documents relating to such reserves and any footnotes.

The subpoena sought all documents in the actual or constructive possession, custody, or control of Textron or its accountants. The district court held oral arguments where the government reiterated that it was seeking tax accrual documents prepared by Textron or E&Y. Textron argued that it had created the documents in anticipation of a dispute with the IRS regarding its tax returns.

1 Textron was joined by amici curiae Committee on Taxation and Committee on Corporate Reporting of Financial Executives International, Chamber of Commerce of the United States of America, and Association of Corporate Counsel. The IRS has submitted two scholarly articles which we have considered: Dennis J. Ventry, Jr., Protecting Abusive Tax Avoidance, 120 Tax Notes 857 (2008); and Claudine Pease-Wingenter, The Application of the Attorney-Client Privilege to Tax Accrual Workpapers: The Real Legacy of United States v. Textron, 8 Houston Bus. & Tax L.J. 337 (2008).

5

The district court then held an evidentiary hearing on the types of documents included in the definition of "tax accrual workpapers" and the basis for Textron's work product claim. At the evidentiary hearing, the IRS's expert witness, Professor Douglas Carmichael, testified that securities law requires that public companies obtain a letter from an independent auditor approving the company's financial statements, and that part of that audit was an analysis of the company's reserves for covering tax loss. Textron's former Director of Tax Reporting, Roxanne Cassidy, countered that the tax accrual workpapers were created "to determine whether Textron was adequately reserved with respect to any potential disputes or litigations that would happen in the future." Cassidy and Norman Richter, Textron tax counsel, explained that the tax accrual workpapers listed positions Textron was taking on its tax returns that might require that a reserve be set aside. These positions were then analyzed by Textron attorneys who estimated a percentage likelihood that the position would not prevail if challenged by the IRS. Textron calls this the "hazards of litigation percentage." The reserve requirement was calculated by multiplying this percentage times the tax benefit claimed.

In response, Carmichael, the IRS expert witness, contended that public companies prepare these papers every year to comply with securities law regardless of whether they expect litigation. But, Richter testified that the tax accrual workpapers

6

were prepared under the assumption that issues identified would be challenged by the IRS and would need to be defended. He further testified that if Textron did not anticipate any disputes, the tax accrual workpapers would be blank. The IRS disagreed, arguing that some workpapers would nevertheless be necessary to handle deferred taxes or to justify setting aside no tax reserve.

It was undisputed that the IRS audits every Textron return in multi-year cycles. Testimony also showed that in each audit cycle hundreds of IRS adjustments to Textron's returns were simply accepted by Textron. Where Textron and the IRS do dispute tax liability, their dispute could be resolved through a conference with the audit team, by presentation of arguments to the IRS Office of Appeals, or, ultimately, federal court litigation. In seven of the last eight audit cycles, Textron and the IRS have brought at least one issue to the IRS Office of Appeals. Between 1959 and present, Textron and the IRS have litigated three disputes in federal court. Textron admitted that it expected to concede several issues identified in the workpapers, and that the percentage for those issues was listed as "100%." Textron explained that those items were due to instances where legal developments rendered a prior position undefendable.

IRS agents explained that access to the workpapers would help them navigate Textron's 4000 page tax return which was accompanied by "9 four-drawer file cabinets" of paper. The IRS

7

witnesses explained that the tax accrual workpapers would help the IRS understand the substance of a transaction and could help the IRS identify potential issues with Textron's return. An agent testified that in 2006 the IRS adopted the policy of seeking tax accrual workpapers when the audit team becomes aware (sometime through self-reporting) of certain transactions by the taxpayer which potentially constitute abusive tax shelters.

As to the waiver issue, the IRS's expert, Carmichael, testified that companies create these tax accrual workpapers, knowing that they will be shared with an auditor. Evidence also showed that the auditor's code of ethics requires information be kept confidential unless required to be produced in response to a subpoena or other legal obligation. Mark Weston, a partner at E&Y, submitted an affidavit explaining these confidentiality obligations and further stating that E&Y had not in fact disclosed any information regarding Textron's 2001 tax accrual computations. But the IRS expert opined that auditors ultimately owe an allegiance to the investing public and have legal obligations of disclosure in some instances.

Textron countered that it let E&Y look at its tax accrual workpapers, but did not allow E&Y to retain a copy of them. Debra Raymond, a Textron tax manager, averred that she had reviewed E&Y's files and that E&Y did not in fact retain a copy of Textron's workpapers.

8

The IRS subpoena also requested any tax-accrual documents prepared by E&Y regarding Textron. After the evidentiary hearing, the district court found that the "tax accrual workpapers" at issue did not include facts about the issues they identified, and could be described as follows:

1. A spreadsheet that contains:

(a) lists of items on Textron's tax returns, which, in the opinion of Textron's counsel, involve issues on which the tax laws are unclear, and, therefore, may be challenged by the IRS;

(b) estimates by Textron's counsel expressing, in percentage terms, their judgments regarding Textron's chances of prevailing in any litigation over those issues (the "hazards of litigation percentages"); and (c) the dollar amounts reserved to reflect the possibility that Textron might not prevail in such litigation (the "tax reserve amounts").

2. Backup workpapers consisting of the previous year's spreadsheet and earlier drafts of the spreadsheet together with notes and memoranda written by Textron's in-house tax attorneys reflecting their opinions as to which items should be included on the spreadsheet and the hazard of litigation percentage that should apply to each item.

United States v. Textron Inc., 507 F. Supp. 2d 138, 142-143 (D.R.I. 2007). The government objects on appeal that the district court did not include in this definition any tax accrual workpaper that E&Y prepared in its audit of Textron.

The district court then found that the subpoenaed documents were relevant, requested for a proper purpose, and were privileged attorney-client communications. Id. at 143-48. But,

9

the district court found that the attorney-client privilege was waived by disclosure to E&Y. Id. at 151-52.

As to the work-product protection claim, the district court credited Textron's testimony that its ultimate purpose in preparing the tax accrual workpapers was to ensure that it was "'adequately reserved with respect to any potential disputes or litigation that would happen in the future.'" Id. at 143 (quoting the testimony of Norman Richter). The court found that this desire to ensure adequate reserves was also prompted, in part, by its wish to satisfy independent auditors that Textron had properly reported its financial information. Id. at 150. The court acknowledged the First Circuit's "because of" test for determining whether documents were prepared in anticipation of litigation and recognized that under that test there is no protection for "'documents that are prepared in the ordinary course of business or that would have been created in essentially similar form irrespective of the litigation.'" Id. (quoting Maine v. United States Dep't of the Interior, 298 F.3d 60, 70 (1st Cir. 2002)). The district court then held that the tax accrual workpapers "would not have been prepared at all 'but for' the fact that Textron anticipated the possibility of litigation with the IRS." Id. The district court reasoned that while the papers were used to obtain a favorable opinion letter from E&Y regarding Textron's reserves, there would have been no need for such reserves "if Textron had not anticipated

10

a dispute with the IRS that was likely to result in litigation or some other adversarial proceeding." Id. The IRS appeals the conclusion that the work-product doctrine protects the tax accrual workpapers.

The district court then noted that the standard for waiver of the work-product protection was different than for attorney-client privilege. Id. at 152. The district court concluded that disclosure of Textron's tax accrual workpapers to E&Y did not effect a waiver since disclosure "did not substantially increase the IRS's opportunity to obtain the information contained in them." Id. at 153. The district court based its decision on E&Y's professional confidentiality obligations and the fact that E&Y averred that it had not actually disclosed the information. Id.2 The IRS also appeals this conclusion.

The district court then ruled that the government could not make the showing needed to overcome work-product protection since the government sought Textron's "mental impressions, conclusions, opinions or legal theories." Id. at 154 (internal quotations omitted). The IRS does not appeal this conclusion.

2 The district court also found that E&Y had promised to keep the documents confidential. The IRS contends that this fact is not supported in the record. As explained infra n.9, we need not decide this dispute.

11

II. Discussion

Three issues are thus presented for appeal: (1) whether the work-product doctrine protects Textron's workpapers; (2) whether any such work-product protection was waived through disclosure to E&Y; and (3) whether the district court erred in not considering the IRS's request for E&Y's workpapers.

On these evidentiary issues, we "review district court rulings on questions of law de novo; we review district court fact findings for clear error; and we review 'discretionary judgments such as evidentiary rulings' for abuse of discretion." Cavallaro v. United States, 284 F.3d 236, 245 (1st Cir. 2002) (quoting United States v. Mass. Inst. of Tech. ("MIT"), 129 F.3d 681, 683 (1st Cir. 1997)).

"[T]he party who invokes the privilege bears the burden of establishing that it applies to the communications at issue and that it has not been waived." XYZ Corp. v. United States, 348 F.3d 16, 22 (1st Cir. 2003).

A. Work-Product Protection 1. Applicable law

"[T]he work-product doctrine does apply in tax summons enforcement proceedings." Upjohn Co. v. United States, 449 U.S. 383, 386 (1981). The work-product doctrine protects "documents and tangible things that are prepared in anticipation of litigation or for trial by or for another party or its representative." Fed. R.

12

Civ. P. 26(b)(3)(A). In assessing whether a document was prepared in anticipation of litigation, this circuit uses the "because of" test. Maine, 298 F.3d at 68. Under this test, a document is protected "if, 'in light of the nature of the document and the factual situation in the particular case, the document can be fairly said to have been prepared or obtained because of the prospect of litigation.'" Id. at 70 (quoting United States v. Adlman, 134 F.3d 1194, 1202 (2d Cir. 1998)) (emphasis in original). It is also our law that, "the 'because of' standard does not protect from disclosure 'documents that are prepared in the ordinary course of business or that would have been created in essentially similar form irrespective of the litigation.'" Id. (quoting Adlman, 134 F.3d at 1202). As discussed below, we hold in this case that the presence of a business purpose does not defeat work-product protection.

The work-product doctrine originates in Hickman v.

Taylor, where the Supreme Court laid out the problems with allowing the discovery of work product:

Were such materials open to opposing counsel on mere demand, much of what is now put down in writing would remain unwritten. An attorney's thoughts, heretofore inviolate, would not be his own. Inefficiency, unfairness and sharp practices would inevitably develop in the giving of legal advice and in the preparation of cases for trial. The effect on the legal profession would be demoralizing.

And the interests of the clients and the cause of justice would be poorly served.

13

Hickman v. Taylor, 329 U.S. 495, 511 (1947). The doctrine is rightly seen as a protection for the adversary system, not simply the attorney. Coastal States Gas Corp. v. Department of Energy, 617 F.2d 854, 864 (D.C. Cir. 1980) (noting the doctrine does not protect all documents prepared by a lawyer and instead "focuses on the integrity of the adversary trial process itself" (quotation omitted)).

The Restatement has offered a definition of litigation in the work product context:

"Litigation" includes civil and criminal trial proceedings, as well as adversarial proceedings before an administrative agency, an arbitration panel or a claims commission, and alternative-dispute-resolution proceedings such as mediation or mini-trial.

It also includes a proceeding such as a grand jury or a coroner's inquiry or an investigative legislative hearing. In general, a proceeding is adversarial when evidence or legal argument is presented by parties contending against each other with respect to legally significant factual issues.

Restatement (Third) of the Law Governing Lawyers § 87 cmt. h (2000). "'Adversarialness' is the touchstone of this approach to the 'litigation' question . . . ." In re Grand Jury Subpoena, 220 F.R.D. 130, 147 (D. Mass. 2004).

2. Are tax disputes "litigation?"

The IRS argues that preparation of tax returns is not meant to be an adversary process, but rather is a self-reporting regime that relies on the good faith of taxpayers. The IRS reasons

14

that it is not seeking to gain unfair litigation advantage but rather to verify self-assessment in an environment where the taxpayer holds all the relevant information. Textron responds that it routinely engages in administrative disputes and even federal court litigation with the IRS.

Citing Roxworthy, Textron argues that anticipation of audit disputes can constitute anticipation of litigation. United States v. Roxworthy, 457 F.3d 590, 600-01 (6th Cir. 2006) (finding memos prepared analyzing legal arguments supporting and opposing specific tax positions to be protected work product). But, Roxworthy cites another case, Hodges, for the proposition that "a document prepared 'in anticipation of dealing with the IRS . . . may well have been prepared in anticipation of an administrative dispute and this may constitute litigation within the meaning of Rule 26.'" Id. at 600 (quoting Hodges, Grant & Kaufmann v. IRS, 768 F.2d 719, 719-22 (5th Cir. 1985)) (emphasis added). Hodges, however, was not expressing a holding, but simply remanding to the district court to consider a work-product claim. Considering that Roxworthy dealt with anticipation of litigation with respect to specific transactions, Roxworthy does not establish that all documents prepared analyzing tax returns are protected. Instead, Roxworthy holds that documents created to analyze specific areas of likely dispute may be protected.

15

Nonetheless, after considering the applicable test, we conclude that, while not all "dealing with the IRS" during an audit is "litigation," the resolution of disputes through adversary administrative processes, including proceedings before the IRS Appeals Board, meets the definition of litigation. The IRS is correct that preparation and filing of returns relies on good faith self-reporting. But, good-faith disputes regarding the proper application of tax law also arise during the audit process. Thus, though the initial processing of these disputes in the audit process may not be adversarial, the disputes themselves are essentially adversarial; the subject of these disputes will become the subject of litigation unless the dispute is resolved.3

3. Were the tax accrual workpapers prepared in anticipation of litigation?

In addition to the question of whether audit disputes with the IRS are litigation, we must confront the question of whether Textron's tax accrual workpapers are prepared in anticipation of such disputes. The district court reasoned that the business need to cover potential liabilities arose out of the anticipation of potential disputes, and thus would not occur "but

3 Thus, we do not hold that filing tax returns or participating in audits is itself adversarial. Our holding does not and should not change the essentiality of good faith self-assessment. The IRS retains many enforcement tools to ensure compliance with that system. Thus, we do not share in the IRS's negative conclusions about the policy effects of our judgment, which concerns tax accrual workpapers, documents which the IRS admits it seeks only in limited circumstances.

16

for" those disputes. Textron, 507 F. Supp. 2d at 150. The IRS argues that the district court's conclusion is factually and legally wrong.

a. Analysis

The IRS argues that, as a legal matter, the documents were not prepared in anticipation of litigation and that the district court misapplied the "because of" test when it reasoned that the documents would not have been created "but for" the prospect of litigation.

We agree with the district court's finding that one of the purposes behind the creation of the documents was anticipation of litigation. As the district court explained, the need to estimate the likelihood of success in litigation was a result of the need to set up a reserve fund to cover tax positions for which Textron could foresee disputes with the IRS.

Of course, simple relation to litigation is insufficient to trigger work-product protection. Maine, 298 F.3d at 69. Further, work-product protection must be assessed in reference to a "the function that the document serves," not its content. Roxworthy, 457 F.3d at 595. And it is not enough that the preparer "had the prospect of litigation in mind when it" created the documents. Adlman, 134 F.3d at 1204. Rather, the "because of" test "really turns on whether it would have been prepared irrespective of the expected litigation with the IRS." Id.

17

But, here, the function of the documents was to analyze litigation for the purpose of creating and auditing a reserve fund. It can be fairly said that "'the driving force behind the preparation'" of the documents, Roxworthy, 457 F.3d at 595 (quoting Nat'l Union Fire Ins. Co. v. Murray Sheet Metal Co., 967 F.2d 980, 984 (4th Cir. 1992)), was the need to reserve money in anticipation of disputes with the IRS. The district court so found, and we find no clear error. See Cavallaro, 284 F.3d at 245 (clear error standard applied to findings of fact of district court in privilege case).4 In this way, we read the district court's "but for"

4 The dissent challenges this factual conclusion of the district court, but utterly fails to explain why the district court's acceptance of the testimony of Textron's witnesses was clear error. Specifically, the district court found that Textron would have no need to compute reserves if it had not anticipated litigation. Textron, 507 F. Supp. 2d at 150 (recognizing that though "it may be accurate to say that the workpapers helped Textron determine what amount should be reserved to cover any potential tax liabilities," anticipation of disputes drove that very need). Contrary to the dissent's intimations, the district court did find that Textron's tax accrual workpapers were prepared with dual purposes in mind. Id. at 143 (finding Textron's "ultimate purpose" in preparing the workpapers was to ensure Textron was adequately reserved with respect to disputes or potential litigation, but acknowledging that the desire to establish adequate reserves "also was prompted, in part, by its wish to satisfy an independent auditor" and obtain a clean opinion). These findings were based on testimony from Textron's witnesses. The testimony of the government's expert, cited by the dissent, does not contradict the conclusion that Textron was motivated by dual purposes, especially considering that it is not clear that the expert is using the same definition of "anticipation of litigation" that we adopt. See infra. Thus, the district court's fact finding was not clearly erroneous. Rather, it is accurate to say that Textron's tax accrual workpapers would not have been prepared in a similar form irrespective of the potential for litigation.

18

reasoning as a way of expressing its conclusion that the documents would not have been prepared irrespective of the prospect of litigation, but rather were prepared "because of" the risk of disputes and litigation which gave rise to a need to compute and report tax reserves.5 The IRS next argues that the district court found that the "workpapers helped Textron determine what amount should be reserved to cover any potential tax liabilities and that the workpapers were useful in obtaining a 'clean' opinion from E&Y regarding the adequacy of the reserve amount." Textron, 507 F. Supp. 2d at 150. The IRS argues that this finding legally compels a ruling that the tax accrual workpapers are not protected, since documents prepared in the ordinary course of business receive no protection. Maine, 298 F.3d at 70. The IRS further reasons that since the documents were required to comply with reporting and securities obligations, the tax accrual workpapers were prepared pursuant to "regulatory requirements" and are therefore not protected. See Nat'l Union, 967 F.2d at 984. The IRS reasons that, as a simple matter of law, Rule 26 does not apply to

5 The IRS also argues that some form of workpaper would have been prepared even if no disputes were anticipated in order to justify setting aside no reserve. But that is not the context in which the workpapers at issue in this case were created. Rather, the evidence shows that these workpapers project the risks of litigation over specific tax positions and were created in the context of anticipating disputes (and therefore possible litigation) over those positions.

19

"[m]aterials assembled in the ordinary course of business, or pursuant to public requirements unrelated to litigation, or for other nonlitigation purposes." Fed. R. Civ. P. 26 Notes of Advisory Committee on 1970 amendments.

We reject the IRS's contention that the mere presence of a business or regulatory purpose defeats work-product protection. To be sure, "there is no work-product immunity for documents prepared in the regular course of business rather than for purposes of the litigation," even though "litigation is already in prospect." 8 Charles Alan Wright, Arthur R. Miller & Richard L. Marcus, Federal Practice and Procedure, § 2024 (2008) (emphasis added). But, it is also true that "'[d]ual purpose' documents created because of the prospect of litigation are protected even though they were also prepared for a business purpose." Id.; see also In re Grand Jury Subpoena, 357 F.3d 900, 907 (9th Cir. 2004) (adopting Wright and Miller's "because of" test in order to handle "dual purpose" documents).

Thus, the best reading of the advisory committee's note is simply that preparation for business or for public requirements is preparation for a nonlitigation purpose insufficient in itself to warrant protection. The note states that there is no protection for documents created for business, regulatory, or "other nonlitigation purposes." This language suggests the note is considering business and regulatory purposes as nonlitigation

20

purposes, but does not suggest that the presence of such a purpose should somehow override a litigation purpose, should one exist. To the contrary, the language of the note does not suggest that the advisory committee was considering the problem of dual purpose documents. Thus, we do not read the note as stating that there should be no protection where a document is prepared because of the possibility of litigation but also for a business or regulatory purpose.6 Like in the case where a company analyzes pending litigation for the purpose of setting aside a reserve, cf. Adlman,

6 One district court in this circuit has reached an opposite conclusion. In re Raytheon Sec. Litig., 218 F.R.D. 354, 359 (D. Mass. 2003) ("However, even when documents were created 'because of' litigation, documents that are required to be prepared to comply with the law may not be protected."). This approach refuses to protect documents prepared because of the possibility of litigation simply because another possible purpose is present. As explained infra, such an approach ignores the primary distinction between the "because of" and the "primary purpose" test.

The IRS also points to another decision in this circuit where the district court analyzed opinion letters regarding investment vehicles alleged to be tax shelters. Fidelity Intern. Currency Advisor A Fund, L.L.C., ex rel. Tax Matters Partner v. United States, No. 05-40151, 2008 WL 4809032, at *13 (D. Mass. April 18, 2008). The court found those letters' "purpose was not to analyze pending or imminent claims; instead, it was to induce the taxpayers to invest in the strategy (or, put differently, to provide comfort to those investors, particularly with respect to the possible imposition of future tax penalties)." Id. But that court did not hold "dual purpose" documents unprotected. Rather, it essentially made a factual call, which we do not now judge, that the possibility of litigation was too remote. Id. As explained, infra Section II.A.4, in the present case the context surrounding the creation of the documents, namely the need to estimate reserve funds, made it reasonable for Textron to foresee and analyze litigation.

21

134 F.3d at 1200 (describing hypothetical examples where work- product protection applies), here the business purpose derives from and is inextricably related to anticipating litigation. That the anticipation of such disputes (and corresponding potential litigation) also triggered certain business and accounting obligations does not bar the protection of the work-product doctrine. See id. at 1202 ("Where a document is created because of the prospect of litigation, analyzing the likely outcome of that litigation, it does not lose protection under this formulation merely because it is created in order to assist with a business decision."); In re Special September 1978 Grand Jury (II), 640 F.2d 49, 61 (7th Cir. 1980) ("We conclude that the materials . . . were indeed prepared in anticipation of litigation, even though they were prepared as well for the filing of the Board of Elections reports."). In arguing against this conclusion, the IRS points to a Third Circuit case where that court instructed the district court on remand to assess whether a tax reserve file was prepared to assist in litigation or to comply with securities requirements. United States v. Rockwell Int'l, 897 F.2d 1255, 1266 (3d Cir. 1990). In the present case, however, the district court has already conducted such analysis and has concluded that the workpapers were created because of both purposes.

The dissent disagrees with this legal conclusion and relies primarily on the language in Maine that "the 'because of'

22

standard does not protect from disclosure 'documents that are prepared in the ordinary course of business or that would have been created in essentially similar form irrespective of the litigation.'" Maine, 298 F.3d at 70 (quoting Adlman, 134 F.3d at 1202). Based on this language, the dissent charges we are going against established circuit precedent. But, for the reasons described above, it is error to read this "ordinary course of business" language as implying that all documents prepared with some business purpose in mind are necessarily unprotected. The court in Maine was not confronted with dual purpose documents like those at issue here. Rather, that court recited the above quoted language in the course of holding that there was insufficient information to see if the documents at issue were prepared for litigation or rather in the ordinary course of business. Id. Thus, Maine's "ordinary course of business" language should not be read as barring protection for dual purpose documents where one business purpose is present. This conclusion is all the more clear when one consider's Adlman's reason for adopting the "because of" test:

The issue is less clear, however, as to documents which, although prepared because of expected litigation, are intended to inform a business decision influenced by the prospects of the litigation. The formulation applied by some courts in determining whether documents are protected by work-product privilege is whether they are prepared "primarily or exclusively to assist in litigation" -- a formulation that would potentially exclude

23

documents containing analysis of expected litigation, if their primary, ultimate, or exclusive purpose is to assist in making the business decision. Others ask whether the documents were prepared "because of" existing or expected litigation -- a formulation that would include such documents, despite the fact that their purpose is not to "assist in"

litigation. Because we believe that protection of documents of this type is more consistent with both the literal terms and the purposes of the Rule, we adopt the latter formulation.

Adlman, 134 F.3d at 1197-98, quoted in part in Maine, 298 F.3d at 68.

Further, a contrary holding would lead to undesirable results in other cases. Consider a document prepared to analyze a specific litigation in order to compute for an auditor how much must be retained in a litigation reserve fund. Were we to adopt the IRS position that documents created to satisfy audit reporting responsibilities were not protected, opposing counsel in the litigation might be able to discover such a memo, effectively disclosing counsel's ultimate mental impression of the case. In fact, in similar circumstances, the Sixth and Second circuits have suggested that memoranda analyzing a transaction or position to be protected by the work-product doctrine. Roxworthy, 457 F.3d at 600; Adlman, 134 F.3d at 1200-04 (remanding for the district court to apply the correct standard).

24

b. Countervailing concerns Having explained why application of the work-product doctrine leads us to the conclusion that Textron's tax accrual workpapers are protected, we pause to address the IRS's arguments against such a conclusion.

i. Arthur Young

The IRS asserts our decision will run afoul of Supreme Court precedent. In Arthur Young, the Court declined to recognize an accountant's work-product doctrine, thus holding that tax accrual workpapers created by an independent auditor were not protected. United States v. Arthur Young & Co., 465 U.S. 805, 815- 21 (1984). The Court recognized that the auditor's papers necessarily included the taxpayer's own thinking about tax "soft spots." Id. at 813. The Court also rejected some of the fairness concerns implicated in that case. Id. at 820 (rejecting argument that "enforcement of an IRS summons for accountants' tax accrual workpapers permits the Government to probe the thought processes of its taxpayer citizens, thereby giving the IRS an unfair advantage in negotiating and litigating tax controversies"). Further, the Court rejected policy concerns that, without a new privilege, companies would be tempted to withhold information from auditors. Id. at 818-19 (finding that "[r]esponsible corporate management would not risk a qualified evaluation of a corporate taxpayer's financial posture to afford cover for questionable positions

25

reflected in a prior tax return" and concluding that "the independent auditor's obligation to serve the public interest assures that the integrity of the securities markets will be preserved, without the need for a work-product immunity for accountants' tax accrual workpapers").

Following this precedent, we have recognized that "the doctrine of construing the [attorney-client] privilege narrowly . . . has particular force in the context of IRS investigations given the 'congressional policy choice in favor of disclosure of all information relevant to a legitimate IRS inquiry.'" Cavallaro, 284 F.3d at 245-46 (quoting Arthur Young, 465 U.S. at 816). But unlike the Court in Arthur Young, we are not now confronted with the question of whether to recognize a new privilege. Here, the doctrinal decision we face is whether to afford protection to documents created because of both business and litigation -- a question not presented in Arthur Young.7 Since this question has

7 The IRS also suggests that the Supreme Court was comfortable with discovery of taxpayer tax accrual workpapers because it "affirmatively encouraged the IRS's seeking tax-accrual workpapers from the company prior to pursuing such workpapers from the auditor." IRS Reply Brief at 22-23 (citing Arthur Young, 465 U.S. at 820-21 & n.17). But the cited provisions of Arthur Young do not consider the question of work-product protection, and do not endorse any particular method, but rather simply take note that the IRS "has demonstrated administrative sensitivity" by adopting a policy to "take all reasonable means to secure the information from the corporation itself before issuing a summons to the independent auditor." Arthur Young, 465 U.S. at 21 n.17. Whether such "reasonable means" can include summons of a taxpayer's own tax accrual workpapers was not considered by the Supreme Court.

26

broader implications, the Supreme Court's policy judgment that a new privilege is not necessary for the accurate preparation of accountants' tax accrual workpapers is not controlling. Further, our prior analysis of this issue does not rely on the policy considerations rejected in Arthur Young. That case is not an obstacle to our conclusion that protection of dual-purpose documents is consistent with the purpose and doctrine of the work- product doctrine.

ii. El Paso

Because of a contrary outcome in a Fifth Circuit case, the IRS argues that our decision will create a circuit split. See United States v. El Paso Co., 682 F.2d 530, 542 (5th Cir. 1982), cert. denied, 466 U.S. 94 (1984). In El Paso, the Fifth Circuit found that a company's own tax accrual workpapers were not protected as they were created for purposes of complying with securities law:

In sum, we believe that the tax pool analysis does not contemplate litigation in the sense required to bring it within the work product doctrine. The tax pool analysis concocts theories about the results of possible litigation; such analyses are not designed to prepare a specific case for trial or negotiation. Their sole function, from all that appears in the record, is to back up a figure on a financial balance sheet. Written ultimately to comply with SEC regulations, the tax pool analysis carries much more the aura of daily business than it does of courtroom combat. We hold, therefore, that the tax pool analysis and backup memoranda are not protected work product materials.

27

Id. at 543-44. However, in El Paso the Fifth Circuit applied a different definition of the work-product doctrine, asking whether the "primary motivating purpose behind the creation of the document was to aid in possible future litigation." Id. at 542 (internal quotation marks omitted). Thus, the result here does not create a circuit split with El Paso, but is merely the application of an existing split in the definition of the "anticipation of litigation." It is precisely in these "dual purpose" situations that the "because of" test used in this circuit distinguishes itself from the "primary purpose" test used in the Fifth Circuit. Maine, 298 F.3d at 68 (citing Adlman for the proposition that the primary purpose test "is at odds with the text and the policies of Rule 26 because nothing in it suggests that documents prepared for dual purposes of litigation and business or agency decisions do not fall within its scope"). Thus, unlike the Fifth Circuit, we need not assess whether the tax accrual workpapers carry more of one aura than another. We need only conclude that since the analysis was undertaken "because of" the anticipation of litigation, work- product protection applies.

iii. Factual argument

Since some tax positions in the accrual papers carry a 100 percent chance of failing, the IRS argues, as a factual matter, that Textron could not have been anticipating litigation. But, Textron explained that those entries were the result of identifying

28

positions that were incorrect as a result of subsequent legal developments. The IRS also argues that most disputes are resolved informally. But the fact that most disputes are resolved before litigation commences does not mean that analysis of those disputes could not have been in anticipation of "the prospect of litigation" over those disputes. To the contrary, assessing the likelihood that a tax position will not withstand scrutiny necessarily entails analysis which anticipates how litigation of that position would be resolved.

4. Has Textron adequately identified the specific litigation for which the workpapers were prepared?

Adlman and Roxworthy, cases we relied on above, are also not directly on point as they involved documents analyzing the consequences of a specific litigation or specific transaction. The IRS argues that this distinction is controlling, that to protect Textron's tax accrual workpapers is in effect to afford a "blanket" protection -- a work-product doctrine that is so broad it will "swallow" the attorney-client privilege. It is true that we have required proponents of a work-product protection claim to "'identify the litigation for which the document was created . . . and explain why the work-product privilege applies to all portions of the document.'" Maine, 298 F.3d at 69 (quoting Church of Scientology Int'l v. United States Dep't of Justice, 30 F.3d 224, 237 (1st Cir. 1994)). These "identification and explanation

29

requirements are not to be given a hypertechnical construction," but "they can neither be brushed aside nor satisfied by vague generalities." Id.

Here, Textron has met its burden. The tax accrual workpapers identify and numerically evaluate a number of tax positions Textron took on its 2001 returns. As described above, Textron has shown that its analysis of each position was prepared by anticipating the possibility of litigation with the IRS arising over a dispute regarding that position. Thus, the tax accrual workpapers identify specific positions, and the litigation projections were created in anticipation of disputes and possible litigation over those positions.8 The IRS again argues that Textron could not have consistently had specific litigation in mind since most tax disputes were resolved before litigation and since Textron could not reasonably intend to litigate every challenged position. But these facts do not change our analysis. As the district court found, the spreadsheet at issue assessed the likelihood of succeeding in litigation over each specific position. Thus, the

8 The IRS posits that some entries in the tax accrual workpapers may refer to reserves held for deferred taxes over which no litigation is to be expected. That some reserves may not be based on disputes does not change our conclusion regarding computation of reserves that are based on potential disputes. Further, the IRS has not requested that we order the production of only the papers regarding deferred tax reserves, which in any event, would not appear to be relevant to assisting the IRS.

30

function served by creating this analysis is to assess what funds must be set aside in anticipation of litigation over each of the positions identified. That Textron might ultimately decide not to dispute a specific position does not contradict the conclusion that when it estimated its litigation success, it was anticipating specific litigation.

Again, we note that a contrary result would lead to an undesirable result. If we were only to afford work product protection over documents of this sort by requiring a showing, as the IRS suggests, that there was some specific quantum of expectation that the position at issue would mature into full- fledged litigation, we would essentially be offering protection only to the cantankerous and combative taxpayer who intends to thoroughly litigate every position.

Precedent does suggest that a party claiming work-product protection must "'have had a subjective belief that litigation was a real possibility, and that belief must have been objectively reasonable.'" Roxworthy, 457 F.3d at 594 (quoting In re Sealed Case, 146 F.3d 881, 888 (D.C. Cir. 1998)). The district court found that, considering its history, it was reasonable for Textron to conclude that litigation with the IRS was likely. We see the matter differently but reach the same result.

In the context of these facts, we do not think that a history of litigation with the IRS should be the deciding factor.

31

A company with a history of challenging the IRS creates these documents "because of" the same reason as another company without such a history: the possibility of a dispute compels them to anticipate litigation so as to prepare and assess their tax reserve funds. In this case, the anticipation of litigation coupled with securities and reporting requirements forced Textron to analyze and project its likelihood of success in litigation. Thus, Textron was effectively forced to operate under the hypothetical belief that litigation would occur. For this reason, regardless of its specific history with the IRS, it was objectively reasonable for Textron to forecast litigation.

The IRS argues that this holding will "immunize nearly every document generated by lawyers because clients can always be said to be aware of possible litigation." The IRS worries that work-product protection would "encompass essentially all legal advice," since any lawyer analyzing the risks inherent in a legal transaction or position will be able to gain work-product protection by saying that the analysis was prepared in anticipation of litigation regarding that transaction or position. We disagree. As we have just explained, in the case of tax accrual workpapers, the dual purposes -- financial reporting and anticipating litigation -- are necessarily intertwined; the function of preparing adequate financial reports inherently requires Textron to anticipate and analyze litigation. The anticipation of litigation

32

may be similarly intertwined in other cases. See, e.g., Delaney, Migdail & Young, Chartered v. IRS, 826 F.2d 124, 127 (D.C. Cir. 1987) (distinguishing case law requiring a "specific claim" to uphold a claim of work-product protection to certain IRS memoranda analyzing potential legal vulnerabilities). But in many cases, there will not be an obvious objectively reasonable need to foresee litigation.9 Of course, each case will depend on its facts. See, e.g., Adlman, 134 F.3d at 1204 (remanding for further fact finding applying the correct standard). But, in the present case, we conclude that Textron's need to consider the possibility of litigation with respect to each disputed position adequately satisfies this requirement.

9 For example, the IRS argues that protecting Textron's tax accrual papers would lead to protection of documents prepared by a bank to estimate loan failure rates. We disagree. A loan may fail for all sorts of reasons. Tax positions would ultimately succeed or fall based on their legal merit, assessed through litigation. In this case, the combination of the possibility of litigation and reporting requirements required Textron to anticipate litigation. Legal analysis of other transactions and positions may not entail such specific consideration of litigation.

The IRS also fears our holding will somehow lead companies to use lawyers to conduct performance reviews, then seek work-product protection for such documents. We think such a hypothetical is distinguishable. In such a case, the employer may be evaluating the employee's performance for business purposes, regardless of the possibility of litigation. Here, Textron evaluates its tax positions for business purposes precisely because of the possibility those positions will be challenged in litigation. Thus, the business context makes it reasonable for Textron to anticipate and foresee litigation. While such a question is a factual matter for another case, it seems less likely that the function of employee evaluations would similarly make it reasonable for an employer to anticipate litigation.

33

B. Waiver of Work-Product Protection Unlike attorney-client privilege, "work product protection is provided against 'adversaries,' so only disclosing material in a way inconsistent with keeping it from an adversary waives work product protection." MIT, 129 F.3d at 687. Specifically, "disclosure to an adversary, real or potential, forfeits work product protection." Id.

We must decide whether Textron waived work-product protection by showing its tax accrual workpapers to E&Y, its independent auditor. A number of district court cases have concluded that disclosure to independent auditors does not waive work-product protection. See, e.g., Regions Fin. Corp. & Subsidiaries v. United States, No. 06-00895, 2008 U.S. Dist. LEXIS 41940, at *27-28 (N.D. Ala. May 8, 2008); Lawrence E. Jaffe Pension Plan v. Household Int'l, Inc., 237 F.R.D. 176, 183 (N.D. Ill. 2006) ("Disclosing documents to an auditor does not substantially increase the opportunity for potential adversaries to obtain the information."); Merrill Lynch & Co. v. Allegheny Energy, Inc., 229 F.R.D. 441, 448 (S.D.N.Y. 2004) ("A business and its auditor can and should be aligned insofar as they both seek to prevent, detect, and root out corporate fraud."). But see Medinol v. Boston Sci. Corp., 214 F.R.D. 113, 116-17 (S.D.N.Y. 2002) ("Boston Scientific and its outside auditor Ernst & Young did not share 'common interests' in litigation, and disclosures to Ernst & Young as

34

independent auditors did not therefore serve the privacy interests that the work product doctrine was intended to protect."). It is undisputed that Textron and E&Y were not actual adversaries. The IRS argues by analogy to MIT that they were potential adversaries. In MIT, we found work-product privilege waived when MIT disclosed documents to a Defense Department auditing agency. 128 F.3d at 687. The audit agency was reviewing MIT's expense submissions. Id. We found MIT and the agency to be potential adversaries as, "MIT doubtless hoped that there would be no actual controversy between it and the Department of Defense, but the potential for dispute and even litigation was certainly there." Id.

The facts here are different. While the Defense Department auditing agency was reviewing expenses to determine if it would challenge those expenses, E&Y was not auditing Textron in order to identify disputes it would have with Textron, but rather, to decide if it could issue a letter certifying Textron's financial statements. This is a cooperative not adversarial relationship. See Jaffe Pension Plan, 237 F.R.D. at 183 ("[T]he fact that an independent auditor must remain independent from the company it audits does not establish that the auditor also has an adversarial relationship with the client as contemplated by the work product doctrine."). While it is possible to imagine circumstances where E&Y's professional obligations could cause E&Y and Textron to come into conflict on some legal question, the IRS can point to no

35

"conceivable scenario in which E & Y would file a lawsuit against [Textron] because of something E & Y learned from [Textron's] disclosures." Regions, 2008 U.S. Dist. LEXIS 41940, at * 27-28 (citing MIT and the district court's opinion in this case to conclude that there was no waiver in similar circumstances). Whereas MIT's disclosure to the Defense Department auditing agency had the potential to directly trigger a dispute with that agency, no such potential adversity existed here.

Disclosure to a conduit to a potential adversary can also waive work-product protection. Raytheon, 218 F.R.D. at 360 (collecting cases). In other words, waiver occurs upon disclosure to a third party that "substantially increased the opportunities for potential adversaries to obtain the information." 8 Wright, Miller, & Marcus, supra, § 2024 (2008). Though E&Y has professional confidentiality obligations limiting disclosure, the IRS argues that it may be required to turn documents over to the SEC and may, in some circumstances, have a duty to disclose information to protect stockholders. See Arthur Young, 465 U.S. at 817-18 ("The independent public accountant performing this special function owes ultimate allegiance to the corporation's creditors and stockholders, as well as to the investing public.").

The IRS also argues that E&Y may be required to disclose this information in response to a valid subpoena. At first glance, this argument seems circular: a subpoena would only be valid if the

36

protection did not survive. But, on further examination, the question is more complicated. The record establishes that E&Y did not retain a copy of Textron's tax accrual workpapers. Thus, at first we would see little likelihood that E&Y would have to disclose those papers. But, E&Y used those papers, together with its own expertise, in preparing its own assessment of Textron's reserve tax liability. Therefore, the only remaining documents which could be subjected to a risk of discovery are E&Y's own assessments, which incorporate Textron's analysis.

Though we hold Textron's tax accrual workpapers are protected, Arthur Young suggests that E&Y's workpapers may be discoverable. 465 U.S. at 815-821.10 In fact, as we find infra, Textron itself may be required to produce E&Y's workpapers on remand.

Since E&Y's workpapers may be discoverable, the question we must ask is whether disclosure of those workpapers substantially increased the risk that the contents of Textron's workpapers would be disclosed to an adversary. As the Supreme Court has recognized, independent auditors "obtain and assess the opinions, speculations, and projections of management with regard to unclear, aggressive, or questionable tax positions that may have been taken on prior tax

10 Though the parties debate the extent of confidentiality agreements and expectations of privacy between Textron and E&Y, we find such expectations irrelevant here, since a bilateral confidentiality agreement regarding Textron's workpapers would not defeat a valid IRS subpoena for E&Y's workpapers.

37

returns." Arthur Young, 465 U.S. at 812. Further the auditor's workpapers "may document the auditor's interviews with corporate personnel." Id. (emphasis added). Thus, disclosure of E&Y's workpapers might reveal Textron's own analysis. On the other hand, E&Y's affiant, Weston, suggested that E&Y "did not rely solely upon the conclusions of [Textron's] tax counsel" but instead "applied its own professional knowledge and judgment." Because the district court did not address the question of whether the IRS can gain discovery of E&Y's tax workpapers through a subpoena to Textron, it made no factual findings regarding the actual contents of E&Y's workpapers or the extent to which disclosure of such workpapers would effectively constitute disclosure of Textron's own assessment. Rather than rely on vague language suggesting what might be in E&Y's workpapers, we think that this is a question the district court must develop on remand through in camera inspection, if feasible, or through testimony.

C. E&Y's Workpapers

The IRS's subpoena also sought tax accrual workpapers prepared by E&Y that were within the possession, custody, or control of Textron or its accountants. The district court did not include such workpapers in its definition of the documents sought and so did not rule on their discoverability. The IRS asserts that such failure to rule is error.

38

Textron argues that this omission is the fault of the IRS. When convening the evidentiary hearing, the district court called on the parties to present evidence on the question of "[w]hat specific types of documents are included in the 'Tax Accrual Workpapers' and 'backup' documentation" sought by the summons. But the evidence at the evidentiary hearing simply concerned Textron's workpapers, and the IRS did not bring up E&Y's workpapers. Textron thus argues that the IRS should not be permitted to now challenge the scope of the district court's construction of the summons. We disagree.

The summons explicitly sought these documents and the record indicates that the government reiterated its request in its briefing and at the initial oral arguments held before the evidentiary hearing was scheduled. The district court was fairly put on notice of this request, and in fact, when discussing the applicability of I.R.C. § 7525, stated at the oral arguments, "I think you've made your case as far as the Ernst & Young papers." That the evidentiary hearing focused on developing evidence regarding the nature of Textron's tax accrual workpapers does not show that the IRS had abandoned its clearly made request for E&Y's workpapers. In other words, since the IRS "squarely and distinctly" requested E&Y's workpapers, and the district court was on notice of the issue, the argument is not waived. Cf. B&T Masonry Constr. Co. v. Pub. Serv. Mut. Ins. Co., 382 F.3d 36, 40

39

(1st Cir. 2004) ("To preserve a point for appeal, some developed argumentation must be put forward in the nisi prius court -- and a veiled reference to a legal theory is not enough to satisfy this requirement.").

Turning to the merits, as noted above, Arthur Young suggests such documents are not protected and Textron has not argued otherwise. Textron also does not meaningfully dispute the government's prima facie showing of relevance. Instead, Textron contends that we may nevertheless affirm the district court because Textron, the only defendant in this action, does not have possession, custody, or control of these documents.

In the normal discovery context, "so long as the party has the legal right or ability to obtain the documents from another source upon demand, that party is deemed to have control." Mercy Catholic Med. Ctr. v. Thompson, 380 F.3d 142, 160 (3d Cir. 2004); see also 8A Wright, Miller, & Marcus, supra, § 2210 ("Inspection can be had if the party to whom the request is made has the legal right to obtain the document, even though in fact it has no copy."). Textron has provided us with no argument why that test should not apply in the IRS subpoena context.

Textron argues that its witness testified that Textron could not access the documents. But Textron mischaracterizes the record. Textron's witness simply testified that E&Y did not give Textron its workpapers, not that Textron could not obtain copies of

40

the documents upon request. Thus Textron has no evidence showing that it did not have the ability to demand the documents. Textron argues that R.I. Gen. Laws § 5-3.1-22 establishes an independent auditor owns its own workpapers. But, again, it does not establish that Textron does not have the ability or contractual right to request the documents.

The IRS has some evidence to suggest that Textron may have had the right to demand E&Y's workpapers. Raymond, a Textron witness, stated that she had reviewed E&Y's files to confirm that E&Y did not retain Textron's workpapers. The IRS argues that this shows Textron has the right to obtain the documents. Textron argues it shows only a right to temporary access. We need not parse this testimony, since even adopting Textron's interpretation, we could not affirm on this basis. Textron has the burden of showing a lack of control and has produced no evidence to meet this burden. United States v. Lawn Builders of New England, 856 F.2d 388, 392 (1st Cir. 1988) ("'Once the district court has reason to believe that the requested documents exist, the burden then shifts to the summonee to show that he is not in possession of them.'" (quoting United States v. Freedom Church, 613 F.2d 316, 322 (1st Cir. 1979))). Thus, we remand for the district court to determine the factual question of whether Textron can obtain E&Y's workpapers.

41

III. Conclusion

For the foregoing reasons, we affirm the district court's determination that Textron's tax accrual workpapers are protected, affirm that Textron and E&Y were not potential adversaries, but vacate the ultimate determination that work-product protection was not waived, and remand for the district court to reassess, in a manner consistent with this opinion, the question of whether disclosure of E&Y's workpapers would reveal the information contained in Textron's own workpapers. On remand the district court should also assess the discoverability of E&Y's workpapers by determining whether Textron has the legal right or ability to obtain these documents.

Affirmed in part, Vacated in part, and Remanded.

No Costs are imposed.

-Concurring and Dissenting Opinion Follows-

42

BOUDIN, Circuit Judge, concurring in part and dissenting in part. The central issue is whether Textron's tax-accrual work papers are protected from IRS summons because they are attorney work product. Any lawyer might be unnerved to find that his potential adversary could obtain the lawyer's own estimate of his chances of success. But under our own precedent and consistent with other circuits, tax-accrual work papers are not protected because they are prepared for reasons independent of the need to prepare for or conduct litigation.

Federal securities laws require publicly traded corporations like Textron to have their financial statements certified by an independent auditor. See 15 U.S.C. §§ 78l, 78m; 17 C.F.R. § 210 et seq. Key to the audit is evaluating "the adequacy and reasonableness of the corporation's reserve account for contingent tax liabilities." United States v. Arthur Young & Co., 465 U.S. 805, 812 (1983). Tax accrual work papers aid by inter alia

pinpoint[ing] the 'soft spots' on a corporation's tax return by highlighting those areas in which the corporate taxpayer has taken a position that may, at some later date, require the payment of additional taxes" [and also include] "an item-by-item analysis of the corporation's potential exposure to additional liability.

Id. at 813.

In the wake of Enron and other corporate scandals, the IRS began to seek taxpayers' tax-accrual work papers for certain

43

transactions "that [are] the same as or substantially similar to one of the types of transactions that the [IRS] . . . has determined to be a tax avoidance transaction." Treas. Reg. § 1.6011-4(b)(2) (2002) (codified at 26 C.F.R. § 1.6011-4(b)(2)). Here, the IRS' request for Textron's tax accrual work papers was made after an IRS team found that Textron had entered into nine separate "Sale-In, Lease-Out" ("SILO") transactions.11 A qualified privilege exists for attorney work product, but the phrase does not mean any and all work by an attorney that happens to refer to litigation; it means work connected to preparing for or managing litigation. Hickman v. Taylor, 329 U.S. 495, 510-12 & n.9 (1947). The attorney-client privilege is not in any way confined to advice given for use in litigation, Mead Data Central, Inc. v. U.S. Dep't. of the Air Force, 566 F.2d 242, 252-53 (D.C. Cir. 1977), but the latter privilege was held to be waived by Textron and is not at issue on this appeal.

In our circuit and in others, documents are said to be created "in anticipation of litigation," and thus eligible for work product protection if, "in light of the nature of the document and the factual situation in the particular case, the document can be

11 SILO transactions are "a combination of a sale of an asset by a tax-exempt entity" followed by a "lease back to the same entity." Shvedov, Tax Implications of SILOs, QTEs, and Other Leasing Transactions with Tax-Exempt Entities 1, CRS Report for Congress (Nov. 30, 2004). These transactions benefit private entities by allowing them to "deduct depreciation" and "interest." Id. at 10.

44

fairly said to have been prepared or obtained because of the prospect of litigation." Maine v. U.S. Dep't. of the Interior, 298 F.3d 60, 68 (1st Cir. 2002) (quoting United States v. Adlman, 134 F.3d 1194, 1202 (2d Cir. 1998)) (internal quotation marks omitted). The "because of" phrase, standing alone, does not answer the question what happens if the document is created to satisfy some requirement imposed independently of the need to litigate, but our precedent squarely answers the question.

We have held that work-product protection does not extend to "documents that are prepared in the ordinary course of business or that would have been created in essentially similar form irrespective of the litigation." Maine, 298 F.3d at 70 (quoting Adlman, 134 F.3d at 1202) (internal quotation marks omitted). This caveat applies "even though litigation is already in prospect," Adlman, 134 F.3d at 1202 (citing 8 Wright, Miller & Cooper, Federal Practice and Procedure § 2024, at 346 (3d ed. 1998)), and even if "the documents aid in the preparation of litigation." Maine, 298 F.3d at 70.

The district court held that the work papers "would not have been prepared 'but for' the fact that Textron anticipated the possibility of litigation with the IRS" and that if "Textron had not anticipated a dispute with the IRS, there would have been no reason for it to establish any reserve or to prepare the work papers used to calculate the reserve." U.S. v. Textron Inc., 507

45

F. Supp. 2d 138, 150 (D.R.I. 2007). But this misreads Maine: Maine holds that it is not the subject matter discussed in the materials that controls but whether documents are prepared "in the ordinary course of business" or were otherwise independently required, which are both the case with tax accrual work papers mandated by accounting requirements.

The decisions in other circuits bear out this legal interpretation. In Adlman itself, the company in question had "the prospect of litigation [with the IRS] in mind" when it asked for the preparation of a memorandum studying the tax implications of a contemplated restructuring; nonetheless, the Second Circuit allowed that if the memorandum would have been prepared in a similar form without expected litigation with the IRS, then it was not created "because of" litigation. Adlman, 134 F.3d at 1203-04; accord United States v. Roxworthy, 457 F.3d 590, 598-99 (6th Cir. 2006).12 Conversely, in cases finding documents to be protected, it is because the documents were created in order to be useful in litigation. See, e.g., Delaney, Migdail & Young, Chartered v. IRS, 826 F.2d 124, 127 (D.C. Cir. 1987) (protection for "attorneys'

12 See also In re Royal Ahold N.V. Sec. & ERISA Litig., 230 F.R.D. 433, 435 (D. Md. 2005) (no protection for notes and memoranda detailing witness interviews where "principal reason" for documents was "to satisfy the requirement of . . . outside accountants"); Fidelity Int'l Currency Advisor A Fund, LLC v. United States, No. 05-40151-FDS, 2008 WL 4809032, at *12-14 (D. Mass. Apr. 18, 2008) (no protection for opinion letters analyzing "likely tax consequences" of certain transactions where purpose was "to induce the taxpayers to invest in the strategy").

46

assessment of . . . legal vulnerabilities in order to make sure it does not miss anything in crafting its legal case") (emphasis added); see also In re Sealed Case, 146 F.3d 881, 885 (D.C. Cir. 1998) (protection for documents to "protect the client from future litigation about a particular transaction").

To the extent the panel majority rests on the district court's legal interpretation, that interpretation is simply at odds with our decision in Maine and with Adlman on which Maine relied. Nor can the district court's outcome be defended by treating the district court's decision as a factual finding that in this instance Textron prepared these documents both to satisfy the accounting requirements and for possible use in litigation. This is not what the district court found and, given the legal rule adopted in Maine and Adlman, it would not matter if it had so found.

The government offered compelling evidence that the sole reason for the creation of Textron's estimates of risk with respect to individual tax positions was to prepare the reserve figures for the company books and statements and to satisfy the auditors that the reserves were adequate. No contrary evidence was offered by Textron. The district court made no contrary finding and any such finding would have been clearly erroneous. The government's expert witness--the former chief auditor of the Public Company Accounting Oversight Board--testified:

47

Q: Does a public company prepare tax accrual work papers only when it anticipates litigation over the issues identified in the work papers?

A: No. They have to prepare them to support the representations in the financial statements.

Q: So if a public company believes the chances of litigation over the issues in the work papers are remote, it would still prepare those work papers?

A: Yes. because they'd have to be able to support that judgment [to the independent auditor].

Q: And would those work papers be prepared in the same form as work papers involving issues for which the company actually anticipates litigation?

A: Certainly, the basics would be the same....

...

Q: Would the tax accrual work papers ever be blank?

A: I can't think of a circumstance where they would, no.

Textron's own witness acknowledged that it would "have to include in its . . . tax accrual work papers any new transactions that the company entered into that year that there might be some tax exposure on" regardless of whether it anticipated litigation with the IRS. Textron offered no testimony purporting to show that its lawyers had any intention of utilizing its assessment in the conduct of any litigation that might ensue (although it did offer speculation as to how such documents might conceivably be helpful), and such a use is not even arguably the driving purpose behind the document's creation.

The only circuit precedent that directly involves tax accrual work papers held that they were not protected attorney work

48

product. United States v. El Paso Co., 682 F.2d 530, 543 (5th Cir. 1982) (noting that in tax accrual work papers legal analysis is only a "means to a business end").13 The Fifth Circuit looked to the primary purpose of the documents in deciding whether they were created "in anticipation of litigation"; but the "because of" test, as construed in Maine, is certainly no more protective here because it denies protection so long as the papers would have been prepared regardless of litigation.

The government's position is also supported by two scholarly articles that analyzed, and roundly condemned, the lower court's decision. Pease-Wingenter, The Application of the Attorney-Client Privilege to Tax Accrual Workpapers: The Real Legacy of United States v. Textron, 8 Hous. Bus. & Tax L.J. 337 (2008); Ventry, Protecting Abusive Tax Avoidance, 120 Tax Notes 857 (2008). Without endorsing everything in the articles, it is fair to say that they make scholarly mince-meat of Textron's position.

In most matters of privilege, there is a certain arbitrariness as to where to draw the line or, to put it differently, even the most sacrosanct of privileges represents a sacrifice of other interests--usually, as Bentham classically

13 El Paso denied protection for the work papers because the court recognized that the company in question was conducting the relevant analysis because of a need to "bring its financial books into conformity with generally accepted auditing principles." Id. at 543. And, it saw the need to satisfy independent auditors as being "compelled by the securities laws." Id.

49

explained, the interest of truth. 8 Wigmore, Evidence § 2291 (McNaughton Rev. 1961). Here, the matter is not open to a panel for re-examination on policy grounds, but no one should think that the policy arguments are all on one side.

Precisely because the work papers in this instance are required by the financial statement obligations and accounting rules, there is little risk of the chilling effect so often paraded in privilege cases. The Supreme Court in rejecting an accountant's privilege for the preparation of tax accrual work papers pointed out that large companies effectively have to comply with the accounting requirements and would do so with or without a privilege for the accountant. Arthur Young, 465 U.S. at 818-19.

And, while it may seem one-sided to give the government Textron's blue print to weaknesses in Textron's tax returns, the return is massive--constituting more than 4000 pages; the government has an important interest in collecting taxes that are owed; and its inquiries into work papers were focused on a specific type of transaction that had been shown to be open to abuse. So context should be kept in mind before shedding too many tears for Textron.

In all events, it is important for us to adhere to the existing rules of the road. Parties can generally adjust their behavior so long as they know what privilege rules apply. Here, the governing rubric in Maine (First Circuit) and Adlman (Second

50

Circuit) and a precise precedent in El Paso (Fifth Circuit) govern this case. An en banc court could change the rule; a panel majority cannot.