PROST, Circuit Judge.
The United States Army Corps of Engineers ("Army") awarded Appellee Turner Construction Company, Inc. ("Turner") a contract for the construction of a government hospital. Turner's two rival bidders for the contract, Appellant B.L. Harbert-Brasfield & Gorrie ("Harbert-Gorrie") and McCarthy/Hunt, each filed a bid protest with the Government Accountability Office ("GAO"). The GAO recommended that the contract be re-procured without Turner's participation. In due course, the Army announced that it would follow the GAO's recommendation and terminated Turner's contract. Turner then filed the present action in the United States Court of Federal Claims to contest the termination and re-procurement. Because we agree with the trial court that the Army's decision to follow the GAO's recommendation was unreasonable, we affirm the judgment of the Court of Federal Claims.
At issue is a potential organizational conflict of interest ("OCI") that arose during the Army's procurement for the construction of a government hospital at Fort Benning, Georgia. The relationship underlying the potential OCI is an attenuated one—intermittent merger discussions that occurred during the procurement process between Turner's design subcontractor and the parent company of the Army's design consultant.
In June 2007, the Army contracted with the joint venture of Hayes, Seay, Mattern & Mattern ("HSMM") and Hellmuth, Obata, and Kassbaum ("HOK") (collectively,
In June 2008, the Army issued the Phase I solicitation for the hospital contract. Four firms responded to the Phase I solicitation in October 2008, and three firms were selected in December 2008 to advance to Phase II: Defendant-Appellee Turner, Plaintiff-Appellant Harbert-Gorrie, and McCarthy/Hunt. Turner's design partner and proposed subcontractor for the project was Ellerbe Becket ("EB"). The work that EB was to perform constituted about 1.5% of Turner's bid.
HSMM/HOK finished preparing the final Phase II technical requirements package by April 2009. The offerors' Phase II proposals were due several months later on July 7, 2009. Shortly thereafter, the Army's thirty-four member technical review board, which included four consultants from HSMM, reviewed the proposals and prepared comments. On August 24, 2009, the Source Selection Authority determined that Turner represented the best value for the Army. The Army awarded Turner a $333,359,000 contract on September 28, 2009.
At various times during the procurement process, Turner's design partner EB was in merger talks with HSMM's parent company AECOM. Specifically, in May 2008, one month before the Phase I solicitation, AECOM and EB executed a confidentiality agreement regarding AECOM's possible acquisition of EB. Some twenty-five to thirty AECOM employees were involved with the due diligence investigation of EB. None of these employees was part of the HSMM/HOK joint venture team assisting the Army with the hospital project.
AECOM was not the only company pursuing the acquisition of EB. By June 2008, four other firms had expressed interest in buying EB, and EB decided to conduct a multi-party auction. Three firms, including AECOM, submitted bidding letters of interest to EB in late August and early September. AECOM's bid was not the highest offer. EB pursued negotiations with AECOM and the highest bidder, but failed to successfully negotiate mutually acceptable terms with either firm. In November 2008, EB terminated negotiations with all bidders.
AECOM's negotiations with EB terminated before HSMM/HOK provided the Army any advice or materials for the Phase II solicitation. In December 2008, HSMM/HOK submitted draft technical provisions for Phase II to the Army. The Army sought comments from all three participating offerors, and HSMM/HOK revised the provisions in response to this input. The final revised Phase II technical provisions, which were twice as long as the draft version, were released in April 2009.
In May 2009, after the Phase II solicitation materials were complete, AECOM reopened
Some signs of a potential OCI arose during the procurement process. AECOM first became aware of a potential conflict on August 7, 2008. On that day, the senior vice president in charge of the HSMM/ HOK hospital design contract attended an industry forum held by the Army about the upcoming hospital project. At the industry forum, the AECOM executive learned for the first time of EB's interest in the project. The executive consulted with his supervisor about the potential for a conflict of interest to arise if AECOM acquired EB, and his supervisor told him that negotiations with EB "had not been productive." A few weeks later, the executive learned that AECOM's negotiations with EB had been suspended, and he concluded that there was no further potential conflict. The AECOM executive did not report the failed negotiations to the Army.
In February 2009, approximately three months after EB terminated its negotiations with AECOM, the AECOM executive and the Army's project manager exchanged emails about whether any potential OCIs existed. The executive inquired internally as to what relations existed between AECOM and the three Phase II offerors and their subcontractors. On February 6, 2009, the executive reported to the Army that only "teaming relationships" existed.
The AECOM executive did not further consider AECOM's relationship with EB until he arrived to attend sessions of the technical review board on July 20, 2009. Four HSMM employees were participating, and all certified that they had no known conflicts of interest. When the executive was asked to certify that he had no known conflicts of interest, he made several telephone calls to determine whether an OCI existed. Because he knew that EB was Turner's subcontractor, he specifically inquired whether AECOM was currently in negotiations with EB. He learned that negotiations had resumed, and he immediately informed the Army's project manager of a potential OCI and asked to meet with the contracting officer ("CO") to discuss the situation.
The AECOM executive met with the CO and the Army's counsel the following day. The executive discussed his concerns that a potential OCI with a subcontractor might arise but did not identify which offeror or subcontractor, due to confidentiality agreements between AECOM and EB. According to the CO, this was the first time that she became aware of any potential OCI. Because the executive was the only AECOM employee at the technical review board meeting who knew of the potential merger, he proposed that he recuse himself from the meeting to avoid any potential conflict. The Army agreed that the executive's recusal would sufficiently prevent any possible conflict of interest.
Shortly after EB's merger with AECOM was publicly announced on October 26, 2009, the two losing offerors filed bid protests before the GAO. Appellant Harbert-Gorrie's protest alleged that Turner/EB had "unequal access" and "biased ground rules" OCIs. Prior to the filing of these bid protests, the CO had not conducted a comprehensive, documented investigation into whether OCIs existed beyond the meeting between the AECOM executive, Army counsel, and the CO. At the inception of the bid protests, the CO submitted a short five-page statement. She supplemented the statement with a 150-page report after conducting a comprehensive investigation, which included gathering forty-two declarations from involved persons and conducting
On February 19, 2010, the GAO issued an eleven-page decision in Harbert-Gorrie's bid protest. The GAO disagreed with the CO's conclusion that no OCI existed prior to award of the hospital contract and sustained Harbert-Gorrie's "unequal access" and "biased ground rules" protests. B.L. Harbert-Brasfield & Gorrie, JV, B-402229, 2010 WL 1041053 (Comp.Gen. Feb. 16, 2010), at 11. Unequal access OCIs can occur when a company has access to nonpublic information in performing a government contract that may give it a competitive advantage in a later competition for a government contract. The GAO's conclusory, three-page analysis of whether an unequal access OCI existed focused on the potential for access to nonpublic information, rather than on whether AECOM or EB personnel actually obtained access to competitively useful, nonpublic information. Id. at 6-9. The GAO presumed that the forty-nine AECOM employees working on the hospital contract used e-mail in conducting their work for the Corps. It noted that, although AECOM affiliates outside of the team working on the hospital contract did not automatically have access rights to data on the servers by virtue of their employment, the GAO saw no evidence anywhere in the record that there were specific efforts to limit access by others to such e-mail. Focusing on the various ways in which AECOM and EB personnel might have had access to each other's information, the GAO concluded: "[W]ith respect to the AECOM employees who worked on the design contract, without credible evidence that AECOM had systems in place to prevent the receipt of competitively useful information by EB, there is no reasonable basis to assume that the information was not made available to EB employees." Id. at 9. In finding that such potential access existed, the GAO noted that an unequal access OCI may be mitigated through the implementation of an effective mitigation plan, but the GAO ultimately determined that AECOM's access restrictions themselves were deficient and that the CO had not been sufficiently involved in crafting and monitoring any mitigation efforts.
The biased ground rules category of OCIs focuses on the concerns that a company may, by participating in the process of setting procurement ground rules, have special knowledge of the agency's future requirements that may skew the competition in its favor. The GAO's two-page analysis concluded that AECOM had special knowledge of the Army's requirements that would have enabled it to give Turner/EB an unfair advantage in the competition. Id. at 9-11. The GAO noted that there was no evidence that the Army closely supervised AECOM, but suggested that even if it had, it would be unreasonable to assume such supervision would prevent AECOM from using its special knowledge to benefit Turner/EB unfairly. Turner argued that at all times during the Phase II solicitation development, AECOM and EB were not in fruitful negotiations, and that the AECOM employees assisting the Army on the procurement had no knowledge of AECOM's interest in EB. In rejecting this argument, the GAO emphasized the lack of process undertaken by AECOM in determining which of its employees had a need to know of the negotiations and how confidentiality was ensured.
Because the GAO concluded that the Army lacked a reasonable basis for its
On March 19, 2010, the Army announced that it would follow the GAO's recommendation. A few days later, the Army notified Turner that it was terminating its contract for the hospital. On March 31, Turner filed a bid protest in the Court of Federal Claims. The Court of Federal Claims ultimately concluded that the Army's decision to follow the GAO's recommendation was arbitrary and capricious. Turner Constr. Co. v. United States, 94 Fed.Cl. 561 (2010) ("Turner I"). The Court of Federal Claims granted Turner injunctive relief, including restoring the hospital contract to Turner. Id. at 585-86.
In granting Turner's Motion for Judgment on the Administrative Record, the Court of Federal Claims concluded that the GAO's recommendations were irrational. See, e.g., id. at 581. The court acknowledged that rational basis review is not a particularly demanding standard, but it nonetheless concluded that the GAO's decision failed to withstand even that level of scrutiny. Id. The Court of Federal Claims noted that while the GAO was tasked with reviewing the Army's decision for reasonableness, the GAO failed to confront the agency's decision in any meaningful way. Id. The court emphasized that the GAO failed to meaningfully consider the CO's detailed factual findings and improperly substituted its own judgment for that of the CO's. It found that the GAO had relied on mere "suspicion or innuendo" rather than identifying "hard facts" showing an appearance of impropriety, in contrast to the CO's thorough investigation and analysis. Id. The Court of Federal Claims further found that the GAO irrationally discounted the CO's post-award investigations and findings. Id. at 574-76.
On August 5, 2010, the Army reinstated Turner's contract in compliance with the Court of Federal Claims' order after explicitly concluding that reinstatement was "advantageous to the Government." 48 C.F.R. § 49.102(d). Harbert-Gorrie timely appealed. We have jurisdiction under 28 U.S.C. § 1295(a)(3).
We review "rulings on motions for judgment on the administrative record de novo . . . and factual findings based on the administrative record for clear error." PAI Corp. v. United States, 614 F.3d 1347, 1351 (Fed.Cir.2010). In a bid protest case, the agency's award must be upheld unless it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." Id. Our inquiry under these facts requires assessing whether the Army's decision to follow the GAO's recommendation "lacked a rational basis." Id. When applying that standard of review in the context of reviewing an agency's decision to follow a GAO recommendation, we have stated that an agency's decision lacks a rational basis if it implements a GAO recommendation that is itself irrational. Centech Grp. v. United States, 554 F.3d 1029, 1039 (Fed.Cir.2009); Honeywell, Inc. v. United States, 870 F.2d 644, 648 (Fed.Cir.1989). Thus, our task is to evaluate the rationality of the GAO decision. It is well settled that COs are given broad discretion in their evaluation of bids. R & W Flammann GmbH v. United States, 339 F.3d 1320, 1322 (Fed.Cir.2003). When an officer's decision is reasonable, neither a court nor the GAO may substitute its judgment for that of the agency. Id.
Harbert-Gorrie alleges that the Court of Federal Claims made three errors that require reversal: (1) the Court of Federal Claims improperly engaged in a de novo review of the GAO's decision rather than giving it proper deference; (2) the Court of Federal Claims erroneously relied on the CO's post-award, post-protest OCI investigation and analysis; and (3) the Court of Federal Claims misapplied the "hard facts" requirement. In addition, Harbert-Gorrie contends that even if the Court of Federal Claims correctly held that the Army's decision to follow the GAO decision was arbitrary and capricious, the Court of Federal Claims exceeded its jurisdiction and committed an error of law by directing the Army to reinstate Turner's contract. We address each argument in turn.
Harbert-Gorrie first accuses the Court of Federal Claims of improperly engaging in de novo review of the GAO decision. This argument is misplaced because the text of the Court of Federal Claims' opinion makes plain that the court applied the proper standard of review.
The Court of Federal Claims correctly articulated the Honeywell standard that applies here and emphasized that it cannot conduct its "own independent de novo assessment." Turner I, 94 Fed.Cl. at 571, 574, 579. The court acknowledged that "`the controlling inquiry [is] whether the GAO's decision was a rational one.'" Id. (quoting Honeywell, 870 F.2d at 647). Recognizing that its review was a deferential one, the Court of Federal Claims nonetheless concluded that the GAO's decision did not meet this standard. Id. at 581. It concluded that the GAO's decision was not rational because its five-page analysis failed to give any deference to the CO's extensive fact-finding and analysis. As discussed above, this court's precedent from Axiom and PAI make clear that the CO enjoys great latitude in handling OCIs. Because the GAO improperly substituted its own judgment for that of the CO, it was the GAO—not the Court of Federal Claims—that failed to apply the proper deference in conducting its review.
The court then applied the correct standard of review, focusing its extensive analysis on whether the GAO's decision to overturn the CO's determination lacked a rational basis. For example, the GAO concluded that HSMM's and EB's interests were "effectively aligned" in August 2008, when AECOM participated in the multiparty confidential auction for EB that ended before HSMM provided the Army with any draft technical provisions. The Court
The court further concluded that the GAO lacked a rational basis for rejecting the CO's biased ground rules determination. Id. at 581. The CO's analysis of this issue "tracked the precise state of negotiations between AECOM and EB, the exact dates upon which critical changes to the RFP occurred, the exact employees that could have known of the merger, and numerous other facts. Using this data, the CO concluded that no OCI existed." Id. at 580. Reiterating that the GAO's task was to review the agency's decision for reasonableness, the Court of Federal Claims noted that the GAO "failed to address this OCI decision; in fact, the GAO decision on a biased ground rules OCI does not even cite the agency decision that it was tasked with reviewing." Id. Instead, "the GAO cites exactly one piece of information [AECOM's contract with the Army] . . . to support its finding that the record `suggests' that AECOM had `special knowledge' that would have given Turner an unfair advantage." Id. Concluding that the GAO's "failure to meaningfully engage with the agency decision dramatically differs from prior GAO decisions," the Court of Federal Claims indicated that the GAO's determination was not based on hard facts but rather was based on "mere suspicion and innuendo." Id. at 580-81.
The Court of Federal Claims likewise concluded that the GAO lacked a rational basis for rejecting the CO's unequal access determination. Id. at 581-83. Again, the court noted that the GAO "failed to cite any hard facts," pointing only to "vague allegations that someone `may have had access' to unidentified information or that someone `was familiar with the details.'" Id. at 582. The court noted that this "lack of concreteness" in the GAO's analysis was a departure from precedent. Id. In addition, the court pointed out that "apart from simply using the phrase `competitively useful,' the GAO cites to no facts to support [its] conclusion that EB had access to anything of competitive worth." Id. In contrast, the court concluded that the CO carefully assessed the information that AECOM may have had access to and determined that this information "not only lacked competitive utility but was also disclosed to all of the offerors." Id.
Accordingly, the Court of Federal Claims applied the proper standard of review and did not conduct a de novo review as alleged by Harbert-Gorrie.
Next, Harbert-Gorrie argues that the Court of Federal Claims erred in considering the CO's post-award investigation and report in determining that the GAO decision was irrational. Harbert-Gorrie believes that because this investigation was not conducted until after the hospital contract was awarded, it should be given no weight because the conclusions and statements therein were dependent on facts not known by the CO during the procurement process. Pointing to the fact that the Army had not conducted a documented investigation prior to the award of the contract, Harbert-Gorrie contends
Contrary to Harbert-Gorrie's contentions, the Court of Federal Claims did not hold that the CO had no duty to evaluate a potential OCI when she first became aware of it. Under FAR § 9.504(a), a CO must "[i]dentify and evaluate potential organizational conflicts of interest as early in the acquisition process as possible" and "[a]void, neutralize, or mitigate significant potential conflicts before contract award." 48 C.F.R. § 9.504(a) (emphasis added). These duties are separate. PAI, 614 F.3d at 1352. Although the FAR requires a contracting officer to identify and evaluate potential conflicts in the early stages of the acquisition process, § 9.504(a) does not require that this preliminary analysis be documented in writing. If the potential conflict is determined to be a significant one, the CO must avoid, neutralize, or mitigate it before the contract award. 48 C.F.R. § 9.504(a). The CO has considerable discretion in determining whether a conflict is significant. PAI, 614 F.3d at 1352. "A significant potential conflict is one which provides the bidding party a substantial and unfair competitive advantage during the procurement process on information or data not necessarily available to other bidders." Id. The FAR therefore requires mitigation of "significant potential conflicts," but does not require mitigation of other types of conflicts, such as apparent or potential non-significant conflicts. Id.
The Court of Federal Claims did not hold that a CO's duty to evaluate potential OCIs is limited to "significant" OCIs or that a CO may ignore and not evaluate known potential OCIs prior to award. Rather, the court held, consistent with PAI, that a CO has two distinct duties under FAR § 9.504(a): the duty to evaluate potential OCIs as early as possible and the duty to mitigate significant potential OCIs before contract award. Turner I, 94 Fed.Cl. at 574-75. The Court of Federal Claims concluded that the CO discharged those duties.
With respect to the timing of the CO's investigations, the court explained that the FAR "does not require a CO, in every single procurement, to review and document whether OCIs exist prior to award." Id. at 575. Courts reviewing bid protests routinely consider post-award OCI analyses and consider evidence developed in response to a bid protest. See, e.g., Masai Techs. Corp. v. United States, 79 Fed.Cl. 433, 449-50 (2007). In addition, as previously conceded by Harbert-Gorrie, "[s]ometimes . . . an OCI cannot be identified until after award" of the contract. Turner Constr. Co. v. United States, 94 Fed.Cl. 586, 591 (2010) ("Turner II"). If the first time an allegation or evidence of a potential OCI appears is after award, then the earliest time to evaluate that potential OCI as countenanced by § 9.504(a)(1) might be at that time. A CO's post-award evaluation can clear the air of any OCI taint by showing that no significant OCI existed. If, however, the CO's post-award evaluation shows that a significant potential OCI did exist and went unmitigated in violation of § 9.504(a)(2), then serious remedial actions are appropriate.
Harbert-Gorrie contends that the Court of Federal Claims erred by not limiting its analysis to assessing the CO's actions in July 2009 when she first became aware of a potential OCI, but that is not the proper inquiry. In this case, the GAO should
Harbert-Gorrie's last attack on the merits of the Court of Federal Claims' decision is based on the court's application of the "hard facts" requirement. Harbert-Gorrie acknowledges that an OCI must be based on "hard facts; a mere inference or suspicion of an actual or apparent conflict is not enough." PAI, 614 F.3d at 1352. It claims that the Court of Federal Claims misapplied this requirement. Specifically, Harbert-Gorrie argues that the Court of Federal Claims' analysis indicates that it required that the "hard facts" show an actual OCI rather than merely the "appearance" of an OCI. These claims are unavailing.
The Court of Federal Claims explicitly acknowledged that "hard facts" do not need to show an actual conflict—a potential conflict can be sufficient. Turner I, 94 Fed.Cl. at 573. Rather, the Court of Federal Claims drew a distinction between circumstances where "hard facts" indicate the existence or potential existence of impropriety and circumstances such as these where the finding of an OCI relies on inferences based upon "suspicion and innuendo." Id. (citing C.A.C.I., Inc. v. United States, 719 F.2d 1567, 1582 (Fed.Cir.1983)). Here, the GAO relied on its inference that some unnamed HSMM employees "may have had access" to unidentified information. The Court of Federal Claims found this inference to be flawed because the GAO only found possible rather than actual access. Further, this possible access was to "unidentified information" rather than specific, sensitive information. The GAO cited to no facts supporting its conclusion that EB had access to any information of competitive worth. The CO specifically identified all information "to which AECOM may have had access" and found that the information lacked competitive utility and was actually disclosed to all of the Phase II offerors. Because an unequal access OCI requires that a firm have access to non-public information that is competitively useful, Axiom, 564 F.3d at 1377 n. 1, the Court of Federal Claims did not err in its application of the hard facts requirement.
Beyond contesting the merits of the Court of Federal Claims' decision, Harbert-Gorrie also contends that the Court of Federal Claims exceeded its "jurisdiction and authority" by directing the Army to reinstate Turner's contract. Harbert-Gorrie asserts that although this case was filed as a bid protest, the termination of Turner's contract is actually not a bid protest claim but is more properly characterized as a claim under the Contract Disputes Act of 1978 ("CDA"). Harbert-Gorrie argues that Turner should therefore have been required to comply with the provisions of the CDA in order to establish jurisdiction in the Court of Federal Claims for the contract termination aspect of Turner's claim. Because Turner did not satisfy the procedural requirements for a CDA claim, Harbert-Gorrie concludes that the Court of Federal Claims did not have jurisdiction to restore the contract. In other words, Harbert-Gorrie contends that even if the Army's decision to follow the GAO recommendation was arbitrary and capricious, the only injunctive remedy that the Court of Federal Claims could have ordered was the inclusion of Turner in the re-solicitation for the hospital contract.
Injunctive relief is appropriate if it "enjoin[s] the illegal action and return[s] the contract award process to the status quo ante." Parcel 49C Ltd. P'ship v. United States, 31 F.3d 1147, 1153 (Fed.Cir. 1994). In this case, the Army had already determined that Turner represented the best value offer and lawfully awarded it the contract. The Army then acted arbitrarily and capriciously by following the GAO's irrational recommendation to overturn that award and re-procure the contract. The Court of Federal Claims fashioned a remedy that restored the parties to their positions before the Army's unlawful action. This remedy included enjoining the re-procurement of the contract and ordering that Turner's contract be reinstated. The Army did not challenge the Court of Federal Claims' ability to order reinstatement as a remedy. Rather, in implementing the Court of Federal Claims' order, the Army examined the circumstances and independently concluded that reinstatement was in fact advantageous to the Army. See 48 C.F.R. § 49.102(d). The fact that the Army has not objected to the scope of the trial court's injunction reinstating the contract is significant, as the government is the only party with standing to object to that relief. Harbert-Gorrie participated in a procurement that has now been held to be lawful, and it has pointed to no legal theory that would give it a right to a second procurement. To the extent that the trial court's order to reinstate Turner's contract is subject to challenge as being beyond the proper scope of the remedy in a bid protest case, that is a matter that could have been raised by the party being enjoined, but not by a third party whose rights are not affected by the reinstatement order. Accordingly, we do not disturb the remedy of reinstatement awarded by the Court of Federal Claims.
The judgment of the Court of Federal Claims is affirmed.
