Lоckheed Martin Federal Healthcare, Inc. (“Lockheed”) and the government (collectively, “Appellants”) appeal the February 26, 2008 decision of the United States Court of Federal Claims setting aside the United States Army’s award of a TRICARE contract to Lockheed, effective July 21, 2008. For the reasons set forth below, we reverse.
I. BACKGROUND
The United States Department of Defense offers a health care program called TRICARE for active and retired members of the military and them families. This program is managed by the TRICARE Management Activity (“TMA”). TMA hires contractors to perform various services relating to TRICARE operations. TMA classifies the services for which it hires contractors into two categories: (1) “purchased care” requirements, which are requirements for actual health care services; and (2) “non-purchased care” requirements, which are requirements for non-health care services such as support services and infrastructure. The contract at issue in this case is for a non-purchased care requirement.
Because TMA hires contractоrs to provide a wide variety of services, there are concerns that in some circumstances contractors that are awarded multiple contracts may be subject to organizational conflicts of interest (“OCIs”). Sections 9.500 et seq. of the Federal Acquisition Regulation (“FAR”) provide rules and procedures for “identifying, evaluating, and resolving organizational conflicts of interest.” 48 C.F.R. § 9.500(a). Under 48 C.F.R. § 9.504(a), contracting officers (“COs”) are required to “analyze planned acquisitions in order to (1) [ijdentify and evaluate potential organizаtional conflicts of interest as early in the acquisition process as possible; and (2) [a]void, neutralize, or mitigate significant potential conflicts before contract award.”
In order to assist COs and contractors in identifying OCIs, TMA classifies its non-purchased care requirements into three categories:
Category 1: TMA Internal Support: Services which, by their very nature, give the Contractor access to extensive data about the contracts of all other TMA contractors.
Category 2: Program Management Siuppori: Services which assist TMA in planning and managing its activities and programs. This includes, for example: requirements analysis, acquisition support, budget planning and management, *1377 business process reengineering, program planning and execution support, and independent technical management support.
Category 3: Product Support: Services or end items required to meet the mission requirements of TMA’s non-purchased care activities and programs. This includes, for example: concept exploration and development; system design; system development and integration; COTS procurement and integration; internal development testing; deрloyment; installation; operations; and maintenance.
Axiom Res. Mgmt., Inc. v. United States,
Contractor participation in more than one of these areas may give rise to an unfair competitive advantage resulting from access to advance acquisition planning, source selection sensitive or proprietary information. Furthermore, contractor participation in more than one area may give rise to a reаl or apparent loss of contractor impartiality and objectivity where its advisory or planning assistance in one area potentially affects its present or future participation in another area.
Id.
The contract at issue is for program management support for the TRICARE Acquisitions Directorate and falls within Category 2. The RFQ was issued on July 30, 2006, and bids were submitted by Loekheed and Axiom Resource Management, Inc. (“Axiom”), the incumbent contractor for the services, on August 14, 2006. After rating each bid, the CO awarded the contract tо Lockheed on September 19, 2006.
On September 25, 2006, Axiom filed its first bid protest with the Government Accountability Office (“GAO”), alleging, inter alia, that the award to Lockheed was illegal because of an unmitigated OCI. The GAO dismissed the protest on October 31, 2006, after the CO notified the GAO that he would analyze the alleged OCI and issue a new source selection decision. Approximately one month later, the CO completed his OCI assessment, concluding that an unequal access to information conflict 1 would arise in the future if Lockheed were to bid on requirements for purchased care services. However, the CO also concluded that the mitigation plan Lockheed submitted along with its bid was sufficient to protect the government against this potential future OCI. Accordingly, the CO awarded the contract to Lockheed.
Upon learning of the renewed award, Axiom reasserted its allegation that Lockheed’s OCI disqualified it from performing the contract in a second bid protest with the GAO. In response, the CO again agreed to reevaluate the alleged OCI. After another review, which included аn extensive analysis by TMA, the CO again concluded that an unequal access to information conflict might arise if Lockheed bid on purchased care requirements in the future. However, the CO also concluded that this OCI and any potential OCIs arising out of Lockheed’s work on Category 3 contracts were resolved by Lockheed’s mitigation plan and a prohibition against Lockheed bidding on future purchased *1378 care requirements. The CO again awarded the contract to Lockheed.
On April 3, 2007, Axiom filed a third bid protest based on the same аlleged OCI. On July 12, 2007, the GAO denied the protest, finding that the CO’s “extensive analysis” and “comprehensive approach used to address[ ] any conflicts that may arise” were not unreasonable.
In re Axiom Res. Mgmt., Inc.,
No. B-298870.3,
On July 17, 2007, Axiom filed a complaint in the United States Court of Federal Claims alleging, inter alia, that the government violated FAR § 9.500 et seq. and acted arbitrarily and capriciously by awarding the contract to Lockheed.
Axiom I,
abused his discretion in violation of FAR § 9.5 by awarding the Task Order to Lockheed Martin, without develoрing a mitigation plan that does not afford Lockheed Martin any significant competitive advantages, is enforceable, i.e., subject to court order, and otherwise does not impose any anticompetitive effects on future competition.
Id. at 600. However, the court did not immediately enter a permanent injunction barring Lockheed from performing the contract. Instead, the court asked for additional briefing from the parties and stated that it would request advice from the Federal Trade Commission Bureau of Compеtition on whether the public interest favored an injunction. Id. at 601.
On February 26, 2008, the court issued its final decision.
Axiom Res. Mgmt., Inc. v. United States,
II. DISCUSSION
Appellants allege that the Court of Federal Claims committed two principal errors. First, they argue that the court violated established principles of administrative law by permitting Axiom to supplement the record with affidavits created for litigation and then extensively relying on those affidavits to support its decision. Second, they assert that the court failed to properly review the record under the “arbitrary and capricious” standard set forth in the Administrative Procedure Act (“APA”). We agree on both grounds.
A. Supplementation of the Record
“Evidentiary determinations by the Court of Federal Claims, including motions to supplement the administrative record, are reviewed for abuse of discretion.”
Murakami v. United States,
*1379
Appellants argue that the Court of Federal Claims erred by permitting Axiom to supplement the record with materials that were not before the agency, including legal pleadings filed before the GAO, declarations of Axiom’s employees, and declarations from consultants retained for litigation. During a telephone conference with the trial court, the government objected to Axiom’s request to add these documents to the record. The government acknowledged during the conference that the law of the Court of Federal Claims allows supplementation in at least some circumstances identified in
Esch v. Yeutter,
Well, let me cut to the chase here. My practice ... since I’ve been on the Court is to allow everybody to put ... whatever they want to put into the record in trial and even in an administrative record to supplement.
My own view is that I don’t know what’s important or not until I finally get around to looking at the record, which would be some time from now.
Yоu know, I think that I have enough experience. Just because it’s in the record doesn’t mean I’m going to rely on it for any reason. I may never even bother to do anything with it, but I do think it’s better to get it in there.
If it goes up on appeal, that way the Appellate Court has a full record to work with, and the parties have put in everything that they feel they need to have to put forward their best argument. I do it on both sides.
I’m not going to change that practice in this case. I don’t see any prejudice to the government because you don’t know what I’m going to do one way or the other with any of the stuff. The same thing for the Plaintiff. If they want to put it in, I’ll put it in.
J.A. 2034. When the government then requested to add its own supplementary evidence to the record, the court urged government counsel not to “hold back,” saying: “I let everybody put in what they want to ... put in. The world will not come to an end. Western civilization will not crumble based upon this value judgment.” J.A. 2035.
While we recognize the need for an adequate record during judicial review, the parties’ ability to supplement the administrative record is limited. In
Camp v. Pitts,
the Supreme Court stated that “the focal point for judicial review should be the administrative record already in existence, not some new record made initially in the reviewing court.”
We conclude that the trial court abused its discretion in this case by adding Axiom’s documents to the record without evaluating whether the record before the agency was sufficient to permit meaningful judicial review. The court made clear that it would freely allow the parties to supplement the rеcord “with whatever they want,” and, by so doing, failed to make the required threshold determination of whether additional evidence was necessary.
In its written decision finding that the government violated FAR § 9.5, the court offered an additional explanation for its determination that supplementation was warranted in this case:
Although the Federal Circuit has not ruled on whether, or the circumstances under which, the Administrative Record may be supplemented, the United States Court of Appeals for the D.C. Circuit has allowed the Administotive Record to be suрplemented in circumstances similar to those presented in this bid protest. See Esch v. Yeutter,876 F.2d 976 , 991 (D.C.Cir.1989) (“[T]he procedural validity of the Department’s action ... remains in serious question. Particularly in this context, it may sometimes be appropriate to resort to extra-record information to enable judicial review to become effective.”).
Axiom I,
Second, and more critically,
Esch’s
vitality even within the D.C. Circuit is questionable in light of more recent opinions by that court which demonstrate a more restrictive approach to extra-record evidence.
See, e.g., IMS, P.C. v. Alvarez,
The focus of judicial review of agency action remains the administrative record, which should be supplemented only if the existing record is insufficient to permit meaningful review consistent with the APA. Faced with the request to supplement the administrative record in this case, the Cоurt of Federal Claims should have determined whether supplementation of the record was necessary in order not “to frustrate effective judicial review.”
Pitts,
B. The Court of Federal Claims’ Review of the Record
Appellants argue that the Court of Federal Claims’ decision to enjoin Lockheed’s performance of the contract resulted from the use of an incorrect standard of review on an improperly supplemented record. Specifically, they allege that the court erred by declining to use the APA’s “arbitrary and capricious” standard and instead undertaking what was essentially a de novo review of the CO’s evaluation of the mitigation efforts. They contend that the record before the CO sufficiently supports the GO’s decision under a proper application of “arbitrary and capricious” review.
1. The Court of Federal Claims’ Standard of Review
Before the trial court, Appellants argued that the court’s task under the APA was to determine whether the CO acted reasonably in awarding the contract to Lockheed. The court, however, disagreed. In the court’s view, “reasonableness” is the proper standard of review “when the court’s evaluation is made under the APA’s ‘arbitrary and capricious’ prong, but [not] where the record contains substantial evidence that one or more FAR provisions have been violated.”
Axiom I,
We agree with Appellants that the court erred by failing to review the CO’s decision under the “arbitrary and capricious” standard set forth in 5 U.S.C. § 706(2)(A). In
Impresa Construzioni Geom. Domenico Garufi v. United States,
this court explained that bid protest cases are reviewed under the standard set forth in the APA.
FAR § 9.504(a) provides that “contracting officers shall analyze planned acquisitions in order to (1) [i]dentify and evaluate
*1382
potential оrganizational conflicts of interest as early in the acquisition process as possible; and (2) [ajvoid, neutralize, or mitigate significant potential conflicts before contract award.” 48 C.F.R. § 9.504(a). Section 9.504(e) further provides that “[tjhe contracting officer shall award the contract to the apparent successful offeror unless a conflict of interest is determined to exist that cannot be avoided or mitigated.”
Id.
§ 9.504(e). However, the FAR recognizes that the identification of OCIs and the evaluation of mitigation proposals are fact-specific inquiries that require the exercise of considerable discretion.
See
48 C.F.R. § 9.505 (“Each individual contracting situation should be examined on the basis of its particular facts and the nature of the proposed contract. The exercise of common sense, good judgment, and sound discretion is required in both the decision on whether a significant potential conflict exists and, if it does, the development of an appropriate means for resolving it.”);
see also ARINC,
In light of the discretion given to COs, we cannot agree with the Court of Federal Claims that the CO in this case “violated” FAR § 9.504 in such a way as to warrant de novo review of “whether or not there may be [a] potential violation of law” and, if so, whether “the mitigation proposal [is] an actual remedy.”
Axiom I,
2. The Court of Federal Claims’ Decision on the Merits
Appellants next argue that the Court of Federal Claims’ determination that Lockheed’s mitigation plan did not adequately resolve the alleged OCI is contrary to the record before the CO. The court decided
that the CO abused his discretion in violation of FAR § 9.5 by awarding the Task Order to Lockheed Martin, without developing a mitigation plan that does not afford Lockheed Martin any significant competitive advantages,' 4 is enforceable, i.e., subject to court order, and otherwise does not impose any anticom-petitive effects on future competition.
Id. at 600. Because the antitrust concerns were originally raised by the trial court and are not relied upon by Axiom in this appeal, we will limit our review to the adequacy and enforceability of the mitigation plan.
In its discussion of the adequacy of Lockheed’s mitigation plan, the Court of Federal Claims relied heavily on two expert witness declarations that the court
*1383
permitted Axiom to add to the record.
See Axiom I,
As discussed previously, supplementation of the administrative record is only appropriate in limited сircumstances. Because we have not been presented with any persuasive explanation of why the record before the CO precluded effective judicial review in this case, we do not find the court’s reliance on the Richards and Adams declarations well placed. However, even if we were to give some weight to these declarations, they do not end the inquiry. After all, a decision is not necessarily unreasonable simply because the disappointed bidder is able to find two witnesses who disagree with it.
Notwithstanding the contrary opinions of Axiom’s declarants, we conclude that the CO’s decision to award the contract to Lockheed was not arbitrary or capricious. Mindful of the confidential nature of Lockheed’s mitigation efforts, we note that the CO and TMA reviewed the OCI Mitigation Plan and comparative analysis submitted by Lockheed and determined that the processes and procedures described therein would be sufficient to mitigate the alleged conflicts. We see nothing unreasonable about that determination. Additionally, TMA barrеd Lockheed from bidding on future purchased care requirements. Finally, we agree with the government that it is reasonable for the CO to defer evaluating certain potential unequal access to information conflicts until Lockheed actually bids on future contracts for which it has obtained non-public information by virtue of its performance of the contract at issue in this case.
See ARINC,
Turning to the enforceability of Lockheed’s mitigation plan, the parties agree that the government would have had legal recourse if Lockheed failed to adhere to the plan. Oral Arg. at 30:00-30:44,
available at
http://oral arguments.cafc.usc-ourts.gov/mp3/2008-5072.mp3. However, the trial court doubted that the CO could be trusted with future enforcemеnt of the plan.
See, e.g., Axiom II,
While we doubt that it would ever be appropriate for the court to interfere with the performance of a contract based solely on its belief or suspicion that the CO cannot be trusted, court interference is certainly inappropriate in this case because the CO did not act arbitrarily оr capriciously in evaluating the mitigation efforts. The Supreme Court has warned against undue judicial interference with the lawful discretion given to agencies.
See Norton v. S. Utah Wilderness Alliance,
III. CONCLUSION
For the foregoing reasons, we reverse the Court of Federal Claims’ decision setting aside the United States Army’s contract award to Lockheed.
REVERSED
Notes
. “An ‘unequal access to information’ conflict occurs when a government contractor has access to non-public information in connection with performance of a government contract that may afford a competitive advantage in subsequent competition for a government contract.”
Axiom I,
. Appellants urge us to distinguish our decision in
Banknote Corp. of America, Inc. v.
*1379
United States,
. In an unpublished opinion, the D.C. Circuit repudiated Esch more directly, noting that the eight exceptions listed were dicta and of limited “probative value.”
Peterson Farms I v. Espy,
. It is unclear whether the court ultimately found that Lockheed's mitigation plan, if actually enforced, would be inadequate to mitigate the alleged unequal access to information conflict. While this portion of
Axiom I
suggests that the court viewed the mitigation plan as insufficient, the court stated in
Axiom II
that it did not "have any problem with the mitigation efforts ... other than they're voluntary.”
Axiom II,