Director, Office of Thrift Supervision v. Vinson & Elkins, LLP

Opinion for the Court filed by Circuit Judge SILBERMAN. SILBERMAN, Circuit Judge: Appellant Office of Thrift Supervision (OTS) asked the district court to enforce administrative subpoenas issued to the law *1306 firm of Vinson & Elkins and two lawyers of that firm. The district court denied enforcement. We affirm. I. Barry Munitz sat on the boards of both the United Savings Association of Texas (Association) and its parent company, United Financial Group. After the Association declared bankruptcy, the Federal Deposit Insurance Corporation interviewed Munitz about his role in the failed savings and loan. C. Michael Buxton and Timothy A. Nelson, two lawyers from Vinson & Elkins, represented Munitz at the interview. Munitz’s testimony, which was neither transcribed nor under oath, was recorded only in the notes taken by Buxton, Nelson, and three FDIC lawyers. Several years later, OTS began formally investigating the Association. 1 OTS deposed Munitz, this time under oath and with a court reporter. One of OTS’ main concerns was the relationship between the Association and its direct and indirect shareholders, including MCO Holdings (later MAXXAM), Federated Development Company, and The Drexel Burnham Lambert Group. Charles Hurwitz was Chairman and CEO of both MCO and Federated, which, combined, owned the largest block of stock in the Association’s parent company. OTS suspected that Hurwitz might have pressured the Association to buy “junk bonds” to benefit his own business interests rather than those of the Association. The following interchange took place between OTS and Munitz: Q: Did Hurwitz ever mention to you that [the Association’s] purchase of junk bonds from Drexel was helpful to MCO’s relationships with Drexel? A: No. Q: Did anyone ever tell you that? A: No one ever told me that. People have asked me that, but no one has ever told me that. Before the deposition, OTS obtained notes of Munitz’s prior testimony from two of the three FDIC lawyers who had been present at the FDIC interview. The FDIC attorney notes suggest that Munitz may have answered the same question somewhat differently in his FDIC interview: Hurwitz Motivations— 1. Peer group pressure — “who was going down & who was still alive.” 2. Did not want to hurt his reputation among peers and Milken [a former officer of Drexel]. 3. Concern about net worth maint. risk— to MAXXAM. 4. [Association] was “ticket to ride” with Drexel. Because of the purported inconsistency, OTS ordered Buxton, Nelson, and Vinson & El-kins to produce all documents relating to the FDIC interview. Buxton, Nelson, and Vinson & Elkins moved to quash the administrative subpoena, claiming the protection of both the attorney-client privilege and the work-product doctrine. In accordance with its internal procedures, OTS designated a Regional Enforcement Counsel to rule on the motion to quash. After reviewing the documents in camera, the Regional Enforcement Counsel determined that the majority of the documents were protected, but ordered Vinson & Elkins to produce portions of the notes Buxton and Nelson took of Munitz’s testimony during the FDIC interview. Vinson & Elkins declined to produce the notes based on the work-produet privilege, and OTS filed a subpoena enforcement action in district court. The district court refused to enforce the subpoena. Appellant argues that the district court erred in characterizing the interview notes as “opinion work product.” In appellant’s view, the district court owed deference to the Regional Enforcement Counsel’s determinations that the notes were “fact work product” and that OTS had demonstrated sufficient need *1307 to obtain them. Appellees counter that, regardless of whether the notes represent facts or opinions, OTS has failed to show sufficient need. II. Appellant’s basic position is that the district court misapplied administrative law in not deferring to OTS’ determination that the notes were only factual work product and that the agency needed the notes. If that were so then we, in turn, would owe no deference to the district court because its error would be one of law. The difficulty with appellant’s theory is that it simply is using the wrong model. 2 This is not a case in which appellees have challenged agency action. The agency instead has come to court in a subpoena enforcement proceeding. To be sure, even in a subpoena enforcement proceeding the district court is sometimes obliged, in effect, to defer to an agency’s determination even if it is only a litigating position. If the dispute turns on the relevance of the information sought by a government agency, we have said that the district court should not reject the agency’s position unless it is “obviously wrong,” and that the burden, as a practical matter, is on the defendant to meet that test. FTC v. Invention Submission Corp., 965 F.2d 1086, 1089 (D.C.Cir.1992) (relying on United States v. Morton Salt Co., 338 U.S. 632, 70 S.Ct. 357, 94 L.Ed. 401 (1950)). 3 We give the agency a wide berth as to relevance because it need establish only that the information is relevant to its investigation not to a hypothetical adjudication, and as we have explained, the boundary of an investigation need only, indeed can only, be defined in general terms. RTC v. Walde, 18 F.3d 943, 946 (D.C.Cir. 1994); Invention Submission Corp., 965 F.2d at 1089. The scope of the investigation, moreover, is very much dependent on the agency’s interpretation and administration of its authorizing substantive legislation concerning which the agency may enjoy interpretative deference. Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). This ease, however, does not turn on the relevance of the information sought, but rather on the question whether under Fed. R.Crv.P. 26(b)(3) the work-product privilege must yield to OTS’ “showing of substantial need of the materials in the preparation of [its] case and that [it] is unable, without undue hardship, to obtain the substantial equivalent of the materials by other means.” Appellant does not claim that the scope of the work-product privilege is somehow reduced if a government agency seeks information. Therefore, it is not apparent why a district judge should afford a government agency’s articulation of its need greater weight than it would a private litigant. III. The parties do not dispute that the interview notes are attorney work product, but they argue at length about whether the notes represent facts or opinions. Appellant argues that the notes are merely an historical account of Munitz’s testimony; appellees maintain that they also reveal the lawyers’ mental impressions and legal theories. A party can discover fact work product upon showing a substantial need for the materials and an undue hardship in acquiring the information any other way. Opinion work product, on the other hand, is virtually undiscoverable. Fed.R.Civ.P. 26(b)(3); Upjohn Co. v. United States, 449 U.S. 383, 401-02, 101 S.Ct. 677, 688-89, 66 L.Ed.2d 584 (1981). Appellees argue, and the district court agreed, that a lawyer’s interview notes are always opinion work product, undiseoverable without “extraordinary justification.” But *1308 this proposition goes too far. We recently observed that under certain circumstances purely factual material embedded in attorney notes may not deserve the super-protection afforded to a lawyer’s mental impressions. See In re Sealed Case, 124 F.3d 230, 236-37 (D.C.Cir.1997). At some point, however, a lawyer’s factual selection reflects his focus; in deciding what to include and what to omit, the lawyer reveals his view of the case. Id. Articulating the degree of selection necessary to transform facts into opinions and the standard of review we should employ of a district court determination are difficult matters that we save for another day. We affirm the district court’s alternate holding, that appellant has not shown sufficient need for the notes under either standard. Appellees argue and we agree that the district court’s assessment of appellant’s “need” and “hardship” is reviewable under the lenient abuse-of-discretion standard. That is so because it calls for a judgment along a range on which reasonable judges could differ. Therefore, the even more deferential standard of review than “clearly erroneous” would normally apply. Appellant urges, however, that in this case de novo review is appropriate, at least insofar as the district court’s judgment is infected by the legal error of declaring that attorney’s notes are necessarily opinion work product, see, e.g., FTC v. Texaco, 555 F.2d 862 (D.C.Cir. 1977) (en banc). But that point does not bear at all on the question of OTS’ need. Appellant also suggests — and this argument seems particularly strained — that it satisfies the need test as a matter of law because the notes would be admissible at trial. The government relies on dicta in a concurring opinion in United States v. Nobles, 422 U.S. 225, 242-54, 95 S.Ct. 2160, 2172-78, 45 L.Ed.2d 141 (1975). Justice White observed that “even in the pretrial discovery area ... work-product notions have been thought insufficient to prevent discovery of evidentiary and impeachment material.” Id. at 249, 95 S.Ct. at 2175. It might be that the party seeking a subpoena in a particular case could show greater need for material that would be introduced into evidence at trial than for material that would provide investigative leads. But we can imagine situations where the purely investigative material would be thought more important. No matter. No court has held that the party seeking the subpoena can brush aside the work-product privilege — even one based only on factual matters — merely by establishing that the material would be admissible in a hypothetical trial. That brings us to appellant’s claimed need. The government explains that it needs the notes to explore the inconsistencies in Munitz’s FDIC and OTS interviews. But two of the three government lawyers present at the FDIC interview have already given their notes to OTS, and therefore the Vinson & Elkins notes would serve only to reinforce supposed inconsistencies about which OTS already knows. The government also tells us that it needs to discover the facts surrounding the debacle of the Association, yet it has not identified any new information that it hopes to find in the Vinson & Elkins notes. Appellant’s need, as we understand its position, reduces to nothing more than its desire to find corroborating evidence. It is the rare case where corroborative evidence can be thought “necessary.” It is often thought cumulative and, as such, a trial judge has wide discretion to exclude it. If corroboration were sufficient need to overcome the work-product protection, what would be left of the second part of the standard, undue hardship? Undue hardship asks whether the moving party can acquire the information any other way; by definition, a party seeking corroborative evidence has already found a way to get the same information. In sum, we can hardly say that the district court’s determination of insufficient need was “arbitrary, fanciful, or clearly unreasonable.” The judgment of the district court is affirmed. Footnotes [1] FDIC and OTS are not interchangeable agencies, but this is a situation where their power overlaps a bit. As receiver, FDIC stands in the shoes of the insolvent institution and can sue officers or directors for damages under state law. 12 U.S.C. § 1821(d)(13)(B)(i) and § 1821(k) (1994). OTS, on the other hand, has independent authority to bring enforcement actions and can require restitution if it proves unjust enrichment or reckless disregard for the law. 12 U.S.C. § 1818(b)(6)(A) (1994). Here, because of the overlap of authority, the two agencies were cooperating, and FDIC let OTS take the lead. [2] For similar reasons the district court properly ignored, despite appellant’s objection, the Regional Enforcement Counsel’s "conclusion’’ after reviewing the notes in camera. The agency's internal procedures in this situation simply do not bear on its case in district court. The district court properly treated OTS’ “ruling” on the work-product doctrine as no more than a litigating position. [3] We have said that a district court’s determination of relevance is reviewable by us under the clearly erroneous test, FTC v. Anderson, 631 F.2d 741, 746 (D.C.Cir.1979) (quoting FTC v. Lonning, 539 F.2d 202, 210 n. 14 (D.C.Cir. 1976)). Applying a deferential standard on appellate review does not mean that the district court has authority to determine relevancy in the first instance. To the contrary, we apply that standard in light of the great deference the district court owes the agency.