SHELL OIL COMPANY and Atlantic Richfield Company, Plaintiffs -Appellees, v. UNITED STATES, Defendant–Appellant.
This case is an appeal from a decision of the Court of Federal Claims requiring the United States to indemnify certain oil companies for environmental cleanup costs under the Comprehensive Environmental Response, Compensation, and Liability Act of 1978, 42 U.S.C. § 9601, et seq. (“CERCLA”). The Court of Federal Claims initially entered judgment in favor of all four plaintiffs in this litigation: Shell Oil Company (“Shell”), Atlantic Richfield Company (“Arco”), Texaco Inc. (“Texaco”), and Union Oil Company of California (“Union Oil”) (collectively, “the Oil Companies”). Upon discovering that his wife had a financial interest in the parent company of Texaco and Union Oil, however, Judge Loren A. Smith: (1) vacated his 2008 and 2009 summary judgment rulings in favor of the Oil Companies; (2) sua sponte severed Texaco and Union Oil from the lawsuit and directed the clerk of court to reassign their claims to a different judge; (3) reinstated his prior summary judgment decisions with respect to Shell and Arco only; and (4) entered final judgment against the government in the total amount of $68,849,505.88. See Shell Oil Co. v. United States, 93 Fed.Cl. 439 (2010) (liability); Shell Oil Co. v. United States, 93 Fed.Cl. 153 (2010) (damages); Shell Oil Co. v. United States, No. 06cv141 (Fed.Cl. Aug. 4, 2010), ECF No. 80 (final judgment).
The government appeals from the Court of Federal Claims' decision entering final judgment in favor of Shell Oil and Arco, and seeks reversal on a number of grounds, including the trial judge's treatment of the discovered financial conflict. Because we find that the presiding judge was required to recuse himself under 28 U.S .C. § 455(b)(4), and that vacatur is appropriate in the circumstances of this case, we vacate the final judgment and the summary judgment orders on which it was premised, and remand with instructions that this case be reassigned to a different judge.
During World War II, the United States entered into contracts with several oil companies for the production of aviation fuel (“avgas”). Production of avgas resulted in an increased production of hazardous waste which was dumped at a waste site in California (referred to as “the McColl site”). Several decades later, after years of litigation in federal district court in California, the Oil Companies were held liable under CERCLA for the costs of cleaning up the waste dumped at the McColl site.
On February 24, 2006, the Oil Companies filed suit in the Court of Federal Claims seeking reimbursement of the CERCLA cleanup costs from the government based on certain language in the avgas contracts. Because there was extensive discovery and the parties entered into comprehensive stipulations of fact in the underlying CERCLA action, the parties agreed that no further factual development was necessary, and the case was litigated on successive summary judgment motions—one as to liability and the other relating to damages. The trial court's decisions on these summary judgment motions are the subject of this appeal.
First, on February 2, 2008, the court granted the Oil Companies' motion for partial summary judgment on liability, finding that the avgas contracts contained “open ended indemnification agreements” and that the “reimbursement clause of the contracts encompasses costs for the CERLCA cleanup as those costs were ‘charges' and ‘by reason of production of the avgas.” Shell Oil Co. v. United States, 80 Fed.Cl. 411, 417–19 (2008). The Oil Companies subsequently moved for summary judgment with respect to the amount of recoverable damages. On March 31, 2009, the court granted, in part, Plaintiffs' motion for summary judgment as to damages and awarded $84,536,763.65 collectively to Plaintiffs. Shell Oil v. United States, 86 Fed.Cl. 470, 475 (2009).
On April 22, 2009, the government filed a motion for reconsideration asking the court to revisit its decision with respect to the remediation of property damage. The court denied the motion on September 28, 2009.
On October 30, 2009, the trial court entered final judgment awarding the sum of $87,344,345.70 to the Oil Companies as follows:
• Shell Oil: $51,166,317.71
• Union Oil: $16,543,019.08
• Arco: $16,543,019.08
• Texaco: $3,091,989.83.
On November 16, 2009,1 the trial judge conducted a telephone conference and informed the parties that, when he was entering final judgment in this case, he reahzed that his wife had inherited 97.59 shares of stock in Chevron Corporation, which is the parent corporation for Texaco and Union Oil. He explained that his wife inherited the shares in 2004, prior to the initiation of this lawsuit. Transcript, Shell Oil v. United States, No. 06cv141 (Fed.Cl. Dec. 7, 2009), ECF No. 60 at 4:9–16. The judge then indicated that he initially did not discover the conflict because he “was particularly focused on Shell” and was looking for “conflict between Chevron and ownership in Shell or Texaco rather being a part of Shell.” Id. at 3:6–10; 5:21–24.
During the telephone conference, the judge did not identify the specific date on which he became aware of the conflict or whether he had sought a formal advisory opinion on the issue. Instead, he indicated that he “consulted with the various technical powers that be in the judiciary and looked at some of the material” and determined the appropriate resolution was to “break Texaco out of this case, vacate all the orders as they relate to Texaco” and have the clerk's office reassign Texaco to a different judge.2 Id. at 3:13–4:1. The trial judge noted, however, that if the parties had a mutually agreeable alternative he would consider it. Both parties asked for more time to analyze the issue.
On December 10, 2009, the government filed a Motion for Relief from Judgment and for Recusal pursuant to Rule 60(b) of the Rules of the Court of Federal Claims. In that motion, the government argued that recusal was mandatory and unwaivable under 28 U.S.C. § 455(b)(4), and that the presiding judge was required to recuse himself from the entire proceeding, not just with respect to individual parties. The government further argued that the court's orders as to Shell Oil and Arco would still have an unfair preclusive effect with respect to the claims by Texaco and Union Oil given the identity of issues between the Oil Companies.
During the course of briefing on the Rule 60(b) motion, the government obtained a copy of the judge's financial disclosure statement for the calendar year 2008. That disclosure report, which the judge signed and certified on May 15, 2009, disclosed an interest in “Chevron Texaco Stock.” See Exhibit–2008 Financial Disclosure Form, Shell Oil, No. 06cv141 (Fed.Cl. Jan. 7, 2010), ECF No. 66–1.
On December 28, 2009, before briefing on the government's Rule 60(b) motion was complete, the government appealed the court's October 30, 2009 final judgment to this court. The government has explained that it filed the appeal at that time to avoid expiration of the time to appeal the court's judgment.
On February 2, 2010, the judge issued an order indicating that, in light of the government's appeal, he no longer had jurisdiction to rule on the government's pending Motion for Relief from Judgment and Recusal. In the order, the trial judge stated that:
the Government could restore jurisdiction by withdrawing its appeal, in which case the Court would vacate its earlier decisions with respect to Texaco and Union Oil. The Government could then appeal the Shell Oil and Atlantic Richfield portions of the decision. The case almost solely focuses on the facts in Shell's case. It could also then re-litigate the Texaco and Union Oil cases before another judge.
Order, Shell Oil, No. 06cv141 (Fed.Cl. Feb. 2, 2010), ECF No. 71 at 2. Accordingly, the government filed a motion in this court to remand.
On May 19, 2010, this court granted the government's motion to remand the case so that the trial court could consider the government's Rule 60(b) motion to vacate and for recusal. In the order, this court stated that, “we deem the better course is to remand so that the trial court may rule in the first instance on the United States' motions.” Shell Oil Co. v. United States, No.2010–5034 (Fed.Cir. May 19, 2010). We further found that the government's request to have the court assign a “new judge to rule on its Rule 60(b) motion, should first be presented to the Court of Federal Claims and thereafter may be raised on appeal from a subsequent ruling, if appropriate.” Id. Accordingly, the case was remanded.
On the same day this court remanded the case, the government filed a motion asking the Chief Judge of the Court of Federal Claims to transfer the case to a different judge. Motion to Reassign Case, Shell Oil, No. 06cv141 (Fed.Cl. May 19, 2010), ECF No. 72. In that motion, the government requested that the case be transferred to a different judge to address the merits of the government's still pending Rule 60(b) motion for relief from judgment and for recusal.
On May 27, 2010, the trial judge issued an order finding a conflict with respect to Texaco and Union Oil, but not with respect to Shell Oil and Arco. Based on that conclusion, the judge ruled that he would allow Shell Oil and Arco to proceed to appeal “in the interest of justice,” sua sponte severed Texaco and Union Oil from the case, and ordered the clerk's office to reassign their claims to another judge.3 In reaching this decision, the court noted that “[i]t would be manifestly unfair to allow a conflict with two of the four parties to delay the decision for the two companies where no conflict of any kind existed or exists.” Order, Shell Oil, No. 06cv141 (Fed.Cl. May 27, 2010), ECF No. 74. Accordingly, the judge denied as moot the government's: (1) Rule 60(b) motion for relief from judgment and for recusal; and (2) motion for the Chief Judge to transfer the case to another judge. The judge then vacated his October 30, 2009 judgment and its 2008 and 2009 summary judgment orders and reissued his prior decisions with respect to Shell and Arco only. Shell Oil Co. v. United States, 93 Fed.Cl. 439 (2010) (Liability); Shell Oil Co. v. United States, 93 Fed.Cl. 153 (2010) (damages).
On August 4, 2010, the judge entered final judgment against the United States in the amount of $68,849,505.88, allocated as follows:
• Shell Oil: $52,027,916.08;
• Arco: $16,821,589.80
The government timely appealed the final judgment to this court on September 17, 2010. As a general rule, a judge's refusal to recuse under § 455 can be reviewed on an appeal from a final judgment. See 13D Charles Alan Wright, Arthur R. Miller & Edward H. Cooper, Fed. Prac. & Proc. § 3553, at 149 (3d ed.2008) (noting that, where a judge refuses to disqualify “it is safe to say—generally, at least—that such an order can be reviewed on appeal from a final judgment in the case”). This is particularly true where, as here, the party asking for recusal timely raised the issue to the court below. We have jurisdiction under 28 U.S.C. § 1295(a)(3).
Standard of review
Given our exclusive jurisdiction over appeals from final judgments of the Court of Federal Claims, we apply the law of this circuit when reviewing the denial of the government's request for recusal. See 28 U.S.C. § 1295(a)(3) (“The United States Court of Appeals for the Federal Circuit shall have exclusive jurisdiction ․ of an appeal from a final decision of the United States Court of Federal Claims.”). Consistent with the vast majority of courts to consider this issue, we review a judge's failure to recuse for an abuse of discretion. See Phonometerics, Inc. v. Westin Hotel Co., 319 F.3d 1328, 1334 (Fed.Cir.2003) (“Under the law of the Eleventh Circuit, the pertinent law on this recusal issue, a district court's refusal to recuse may be raised in an appeal from the final judgment and is reviewed for an abuse of discretion.”) (citations omitted).
The government argues that the judgment of the Court of Federal Claims is tainted because the judge's wife owns shares in the parent company of two parties to the proceeding: Texaco and Union Oil. Specifically, the government argues that the trial judge was required to disqualify himself under 28 U.S.C. § 455(b)(4) from the entire proceeding, and, because he failed to do so, this court should vacate the lower court's judgment and remand the matter with instructions that the case be assigned to a different judge. The government further argues that the judge's decision to sever Texaco and Union Oil while reissuing the decisions with respect to Shell Oil and ARCO “could have preclusive or prejudicial effect upon the remaining matters,” such that the government would be prejudiced by any order other than one which vacates the judgment at issue in this appeal. Appellant's Br. 30.
In response, the Oil Companies argue that: (1) the trial judge complied with § 455(b)(4) when he severed Texaco and Union Oil from this lawsuit and had them transferred to a different judge; and (2) any error in the judge's decision not to disqualify himself would be harmless because this court reviews the trial court's grant of summary judgment de novo.
For the reasons discussed below, we find that the trial judge was required to recuse himself under § 455(b)(4). Given the circumstances, we conclude that vacatur is the appropriate remedy.4
A. 28 USC § 455(b)(4) Mandates Recusal
Section 455 governs the disqualification or recusal of federal judges. Section 455(a) provides that “[a]ny justice, judge, or magistrate judge of the United States shall disqualify himself in any proceeding in which his impartiality might reasonably be questioned.”5 While § 455(a) operates as a “catchall” recusal provision, § 455(b) lists a number of specific circumstances where recusal is required. The subsection at issue here, § 455(b)(4), provides that a judge shall disqualify himself when he “knows that he ․ or his spouse ․ has a financial interest in the subject matter in controversy or in a party to the proceeding.” A “financial interest” is defined as “ownership of a legal or equitable interest, however small.” 28 U.S.C. § 455(d)(4). It is well-established that the ownership of stock constitutes a “financial interest.” The statute defines a “proceeding” to include “pretrial, trial, appellate review, or other stages of litigation.” § 455(d)(1).
Section 455(b)(4) “expressly provides that the judge must know of his or her interest” before he is required to recuse. Liljeberg v. Health Servs. Acquisition Corp., 486 U.S. 847, 859, 108 S.Ct. 2194, 100 L.Ed.2d 855 (1988) (“A careful reading of the respective subsections makes clear that Congress intended to require knowledge under subsection (b)(4) and not to require knowledge under subsection (a).”). This language “embodies an actual knowledge test regarding disqualifying circumstances and provides a bright line as to disqualification based on a known financial interest in a party—i.e., an equity financial interest of any size is disqualifying.” Chase Manhattan Bank v. Affiliated FM Ins. Co., 343 F.3d 120, 127 (2d Cir.2003); see also Union Carbide Corp. v. U.S. Cutting Serv., 782 F.2d 710, 714 (7th Cir.1986) (“Since the statute forbids only the knowing possession of a financial interest, since Judge Getzen-danner relinquished control of the case as soon as she found out about the financial interest, and since she did not resume control until the financial interest was eliminated, at no time was she in literal violation of the statute.”). Pursuant to § 455(c), a judge “should inform himself about his personal and fiduciary financial interests, and make a reasonable effort to inform himself about the personal financial interests of his spouse and minor children residing in his household.” 28 U.S.C. § 455(c).
Recusal under § 455(b) cannot be waived. Indeed, the statute specifically provides that “[n]o justice, judge, or magistrate judge shall accept from the parties to the proceeding a waiver of any ground for disqualification enumerated in subsection (b).” 28 U.S.C. § 455(e). Although recusal under § 455(b)(4) is mandatory, there is a statutory exception where the judge discloses and divests himself of the financial interest. Specifically, § 455(f) provides that, if a judge or spouse “has a financial interest in a party (other than an interest that could be substantially affected by the outcome),” and this interest appears or is discovered “after substantial judicial time has been devoted to the matter,” then “disqualification is not required if the [judge or spouse] divests himself or herself of the interest that provides the grounds for the disqualification.” 28 U.S.C. § 455(f); see also Kidder, Peabody & Co. v. Maxus Energy Corp., 925 F.2d 556, 561 (2d Cir.1991) (finding that § 455(f) applied where the judge immediately disclosed and divested his stock and the parties had devoted nearly three years to the litigation).
The legislative history reveals that § 455(f) was created:
to permit a judge or his or her spouse ․ to resolve the conflict by divesting themselves of the property creating the conflict. Disqualification would continue to be automatic if the interest in the controversy was one that could be substantially affected by the outcome. For example, a significant stockholder in a closely held corporation with close acquaintances would be expected to recuse himself or herself rather than divest.
In re: Certain Underwriter Defendants, 294 F.3d 297, 304 (2d Cir.2002) (citing H.R.Rep. No. 100–889, reprinted in 1988 U.S.C.C.A .N. 6029–30). Where a judge discovers a financial interest and divests in accordance with § 455(f), disqualification under § 455(b)(4) is no longer required. See United States v. Pappert, 1 Fed.Appx. 767, 769 (10th Cir.2001) (finding that the judge's “prompt divestment prevented her from having to recuse” and that once the judge “divested herself of the stock that created the conflict ․ § 455(b)(4) no longer applied”).
Here, there is no evidence that the judge's wife divested the Chevron stock and therefore § 455(f) is inapplicable.6 Because the stock was not divested, recusal is mandatory under § 455(b)(4), and cannot be waived. Given the lack of divestiture, as soon as he discovered the conflict, the trial judge was required to disqualify himself from the entire proceeding and ask the clerk of court to transfer the case to another judge.
Although the judge explained that he did not discover the interest until he was entering final judgment, the record reflects knowledge of his wife's financial interest in Chevron at least as early as May 15, 2009 when he completed his certified Financial Disclosure Report disclosing an interest in “Chevron Texaco Stock.” While the May 15, 2009 disclosure date post-dates the trial judge's February 2, 2008 and March 31, 2009 opinions addressing the oil companies' motions for summary judgment as to liability and damages, it pre-dates his September 28, 2009 decision denying the government's motion for reconsideration with respect to damages, as well as his October 30, 2009 entry of final judgment.7 Accordingly, the judge should have recused himself as of May 15, 2009 and should have taken no further action other than the ministerial act of transferring the case to another judge. See United States v. O'Keefe, 128 F.3d 885, 891 (5th Cir.1997) (noting that, when a judge recuses himself from a case, he “may take no action other than the ministerial acts necessary to transfer the case to another judge”). Neither entering final judgment nor denying the motion for reconsideration can be construed as “ministerial tasks.” See id. (finding that, post-recusal, the district court judge “performed a discretionary act, not a ministerial act” when he ruled on a motion for reconsideration).
The Oil Companies argue that the judge's decision to sever Texaco and Union Oil satisfied § 455(b)(4) because he is no longer presiding over a proceeding in which his spouse has a financial interest. In other words, the Oil Companies suggest that the judge cured the conflict by “divesting” two of the parties from the case. Although a judge can divest a financial interest to avoid recusal, nothing in the statutory text supports the idea that a judge can avoid otherwise mandatory recusal by severing from the proceeding the parties in which his spouse has a financial interest and transferring those parties to a different judge. Because the divestment exception set forth in § 455(f) applies only to divesting a financial interest in a party, and there is no indication that Congress intended to create an exception whereby a judge can sever or “divest” certain parties from the case to resolve a conflict, we find Plaintiffs' argument is not well-taken. This is particularly true where, as here, there is substantial overlap with respect to the issues involved in the remaining parties' claims, and the matters had been considered jointly throughout the proceedings.
Because the judge's wife owns shares in the parent company of Texaco and Union Oil, § 455(b)(4) requires recusal. See Key Pharms., Inc. v. Mylan Labs., Inc., 24 F.Supp.2d 480, 482 n. 2 (W.D.Pa.1998) (noting that, under § 455, “the owner of stock in a parent corporation has a direct legal or equitable interest in a controlled subsidiary and the judge should disqualify himself”). The judge's decision to sua sponte sever Texaco and Union Oil did not satisfy the statutory requirement of disqualifying himself from the entire proceeding. We emphasize that there is neither an allegation nor suggestion that the judge was unduly influenced by his wife's financial interest, which no one argues was substantial. The statute does not require as much-it simply requires recusal whenever financial conflicts of interest exist, regardless of whether those conflicts affect the outcome of the case. Because the judge was required to recuse himself under § 455(b)(4), his failure to do so violated the statute.
B. The Appropriate Remedy
Given our conclusion that the trial judge erred in not recusing himself from the case, we turn now to the appropriate remedy. Plaintiffs argue that any error in the judge's decision not to recuse is harmless because the judgment is subject to de novo review on appeal. In response, the government argues that, under the Oil Companies' theory, “any conflicted court could enter summary judgment and the error would be harmless” given the standard of review on appeal. Appellant's Reply 3.
As the Supreme Court has noted, “[a]lthough § 455 defines the circumstances that mandate disqualification of federal judges, it neither prescribes nor prohibits any particular remedy for a violation of that duty.” Liljeberg, 486 U.S. at 862. Instead, Congress “wisely delegated to the judiciary the task of fashioning the remedies that will best serve the purpose of the legislation.” Id.
In Liljeberg, the Supreme Court affirmed the Fifth Circuit's decision to vacate a district court judge's final judgment where that judge should have disqualified himself under § 455(a) due to an appearance of impropriety. Id. at 852. Although the Court agreed with the Fifth Circuit that vacatur was appropriate under the facts of that particular case, it explained that harmless error analysis can apply to violations of § 455(a). Id. at 862 (“As in other areas of the law, there is surely room for harmless error committed by busy judges who inadvertently overlook a disqualifying circumstance. There need not be a draconian remedy for every violation of § 455(a).”). The Court concluded that, when deciding whether to vacate a judgment for violation of § 455(a), a court should consider: (1) “the risk of injustice to the parties in the particular case”; (2) “the risk that the denial of relief will produce injustice in other cases”; and (3) “the risk of undermining the public's confidence in the judicial process.” Id . at 864.
While the Supreme Court in Liljeberg did not specifically address whether violation of § 455(b) could constitute harmless error, this court and courts in other circuits have indicated that the “differences between sections 455(a) and 455(b) d[o] not ‘preclude the application of harmless error analysis in the context of a 455(b) violation.’ “ Polaroid Corp. v. Eastman Kodak, Co., 867 F.2d 1415, 1421 (Fed.Cir.1989) (citing Parker v. Connors Steel Co., 855 F.2d 1510, 1527–28 (11th Cir.1988)); see also Patterson v. Mobil Oil Corp., 335 F.3d 476, 485 (5th Cir.2003) (“[W]e are confident that § 455(b) violations are also subject to the doctrine of harmless error.”) (citation omitted).
In Polaroid, this court rejected Kodak's attempt to distinguish Liljeberg on grounds that it dealt with § 455(a) rather than § 455(b). Polaroid, 867 F.2d at 1420.8 Specifically, the panel stated that the “argument is meritless, the Supreme Court having made no such distinction in outlining the test to be used in deciding whether to vacate ‘for violation of § 455.’ “ Id. at 1420–21. The court, citing with approval the Eleventh Circuit's decision in Parker, concluded that the Supreme Court's Liljeberg decision did not draw any distinctions between sections 455(a) and 455(b) with respect to the appropriate remedy. Id. at 1421 (citing Parker, 855 F.2d at 1527–28). Applying the Supreme Court's guidance in Liljeberg, we found that “the risk of injustice to the parties weighs far more heavily on Polaroid's side of the scales” and that there was “little or no risk of injustice in other cases” because “the present denial rest[ed] on the specific facts of this case.” Id. at 1420. As such, this court affirmed the district court's denial of the motion to vacate.
In light of our decision in Polaroid, violations of § 455(b) can constitute harmless error. Accordingly, although violation of § 455(b)(4) mandates recusal, mandatory recusal does not require mandatory vacatur. Instead, circuit courts have discretion with respect to the appropriate remedy for a violation of § 455. For the reasons discussed below, we conclude that, under the circumstances of this case, vacatur is the appropriate remedy.
Applying the harmless error factors articulated in Liljeberg, we first consider the risk of injustice to the parties. See Liljeberg, 486 U.S. at 864. Here, the trial court's decision to reinstate its judgment with respect to Shell Oil and Arco may have a prejudicial effect on the now-severed litigation of Texaco and Union Oil's claims. Indeed, given the identity of issues involved, the parties do not dispute that a decision in this case will control the outcome in the severed case involving Texaco and Union Oil. See Oral Argument at 16:20 (noting that the Texaco case is stayed pending resolution of this case, and that this appeal “will be dispositive in that case, most probably”). The government argues that, if judge's opinions with respect to Shell Oil and Arco are allowed to stand, the government would be precluded from challenging the court's determinations under the doctrine of collateral estoppel because the issues would have already been decided adversely to the government. We agree. Because the Oil Companies' avgas contracts contain substantially similar language and the facts relating to dumping waste at the McColl site are nearly identical as to all four companies, the judgment here could have a preclusive or prejudicial effect in the severed case.
In addition to the risk of injustice to the parties, we find that, under these circumstances, there is also a “risk of undermining the public's confidence in the judicial process.” See Liljeberg, 486 U.S. at 864. This is particularly true given that, several months after disclosing a financial interest in “Chevron Texaco Stock,” the trial judge: (1) denied the government's motion for reconsideration; and (2) entered final judgment in excess of $86 million in favor of the Oil Companies.
The Oil Companies argue that the risks of injustice are non-existent because this court will subject the district court's judgment to de novo review. In support of this position, Plaintiffs assert that “[e]very federal circuit court of appeals to consider the question has held that the availability of de novo appellate review renders harmless any error in failing to recuse under Section 455.” Appellees' Br. 57; see Williamson v. Indiana Univ., 345 F.3d 459, 464–65 (7th Cir.2003) (“On appeal, this court reviews the grant of summary judgment de novo ․ and therefore Williamson has received a full review by an impartial panel”); Higganbotham v. Oklahoma, 328 F.3d 638, 646 (10th Cir.2003) (noting that the court “independently reviewed th[e] issues de novo and concluded that the plaintiff's complaint was properly dismissed” and that any “error by the district court judge in not recusing himself would have been harmless under the circumstances of this case”); Patterson, 335 F.3d at 485 (“Because we review a summary judgment ruling de novo, using the same standards as the district court, the parties are guaranteed a fair, impartial review of the merits of the ruling.”). Unlike the circumstances here, however, the courts in the cases upon which Plaintiffs rely specifically found that the risks of injustice and concern for public confidence set forth in Liljeberg were not implicated. See Williamson, 345 F.3d at 465 (“Likewise, the risk of injustice in other cases and concern for public confidence are not implicated by these facts.”); Higganbotham, 328 F.3d at 646 (considering the risks from Liljeberg and concluding that “none of these risks is present”); Patterson, 335 F.3d at 486 (considering the risk of injustice and the public's confidence in the judicial process in finding that the § 455(a) violation constituted harmless error).
The fact that this court reviews the grant of summary judgment de novo does not supplant our consideration of potential injustice stemming from a trial judge's failure to recuse in accordance with § 455(b)(4). In other words, a judge's failure to recuse does not automatically constitute harmless error whenever there is de novo review on appeal. Otherwise, as the government notes, under Plaintiffs' theory, there would be no need for a trial judge to recuse from any action in which the appellate court will apply de novo review. If that were the case, a judge's financial interest in a party would always be harmless so long as the case is decided on summary judgment. Such a result is untenable. Because we find that the trial judge's failure to recuse in this case was not harmless error, particularly given the risk of injustice and risk of undermining the public's confidence in the judicial process, we conclude that the appropriate remedy is to vacate the district court's orders and remand the case.
For the foregoing reasons, we vacate Judge Smith's final judgment in favor of Shell Oil and Arco, as well as the summary judgment orders on which it was premised. Specifically, we vacate the following: (1) the May 27, 2010 Opinion and Order granting Plaintiffs' cross-motion for summary judgment as to liability (ECF No. 75); (2) the May 27, 2010 Opinion and Order granting Plaintiffs' motion for summary judgment as to damages (ECF No. 76); (3) the Final Order dated August 2, 2010 (ECF No. 79); and (4) the Judgment dated August 4, 2010 (ECF No. 80). This case is hereby remanded with instructions that it be reassigned to a different judge on the Court of Federal Claims.
VACATED AND REMANDED
1. November 16, 2009 was the deadline for the parties to seek rehearing under Rule 59 of the Rules of the United States Court of Federal Claims.
2. Although the judge referred only to Texaco during the telephone conference, it is undisputed that Chevron is also the parent company of plaintiff Union Oil.
3. At this stage, Texaco and Union Oil's claims are pending before a different judge on the Court of Federal Claims in Texaco, Inc. v. United States, Case No. 06cvl411. On July 12, 2010, the Texaco court issued an order staying the case pending resolution of this appeal.
4. The government previously moved for summary reversal on the conflict issue. This court denied that motion, noting that “the better course is to deny the United States' motions without prejudice to the United States raising its issues in its briefs.” Order on Motion, Shell Oil Co. v. United States, No.2010–5161 (Fed.Cir. Jan. 4, 2011). We have noted that “summary disposition is appropriate, inter alia, when the position of one party is so clearly correct as a matter of law that no substantial question regarding the outcome of the appeal exists.” Joshua v. United States, 17 F.3d 378, 380 (Fed.Cir.1994) (citation omitted). Given the unique posture and circumstances of this case, particularly the judge's decision to sever the parties with whom his wife has a financial interest, summary disposition was not appropriate here.
5. Given that section 455 applies to “any justice, judge, or magistrate judge of the United States,” it “includes judges who sit on the United States Court of Federal Claims.” Addams–Moore v. United States, 79 Fed.Cl. 578, 580 n. 1 (2007).
6. During oral argument, counsel for the government stated that: “We are unaware of any divestiture. There was not any divestiture as of the time we filed our notice of appeal or as of the time of severance.” Oral Argument at 13:50, available at http://www.cafc.uscourts.gov/oral–argument–recordings/2010–5161/all.
7. Because the financial disclosures must be filed yearly, the same stock ownership should have been disclosed on each financial disclosure which post-dated the 2004 inheritance. It is unclear from the record whether these disclosures were made, however.
8. The facts in Polaroid are distinguishable from those at issue here. In Polaroid: (1) the trial judge informed the parties at the outset of trial that her mother-in-law owned stock in Kodak but that she did not think it was a disqualifying conflict; (2) at the time of final judgment, neither the judge nor her spouse had a financial interest in a party to the proceeding (and thus § 455(b)(4) did not apply); (3) years later, after the judgment was appealed, this court affirmed, and the case was returned to the district court for assessment of damages and trial on remaining claims, the judge held a pretrial conference explaining that her mother-in-law died and that “she was sua sponte disqualifying herself because she was a legatee and her husband was executor of the estate”; and (4) Kodak sought to vacate all of the judge's orders after six and a half years of litigation. 867 F.2d at 1416–17.
O'MALLEY, Circuit Judge.