897 F.3d 340

D.C. Cir.

2018

UNITED STATES OF AMERICA, APPELLEE v. PAUL JOHN MANAFORT, JR., APPELLANT

No. 18-3037

United States Court of Appeals FOR THE DISTRICT OF COLUMBIA CIRCUIT

Decided July 31, 2018

Before: TATEL, GRIFFITH and WILKINS, Circuit Judges.

On Appeal of a Pretrial Detention Order (No. 1:17-cr-00201-1)

Frank P. Cihlar, Richard W. Westling, and Thomas E. Zehnle were on appellant‘s Memorandum of Law and Fact.

Andrew Weissman and Scott A.C. Meisler, Attorneys, U.S. Department of Justice, were on appellee‘s Memorandum of Law and Fact. Michael R. Dreeben and Adam C. Jed, Attorneys, U.S. Department of Justice, entered appearances.

Opinion for the Court filed by Circuit Judge WILKINS.

WILKINS, Circuit Judge: We consider an appeal of a pretrial detention order issued after revoking release due to the defendant‘s alleged commission of new crimes of witness tampering while released. We affirm.

I.

On October 27, 2017, the United States, by and through a Grand Jury convened by Special Counsel Robert S. Mueller, III, indicted Appellant Paul Manafort on nine criminal charges, including conspiracy against the United States, money laundering, failure to file reports of foreign bank and financial accounts, acting as an unregistered agent of a foreign principal in violation of the Foreign Asset Registration Act (“FARA“), and making false and misleading statements. See Indictment, United States v. Manafort, No. 17-cr-201 (D.D.C. Oct. 30, 2017), ECF No. 13. Following his arrest and arraignment, the District Court released Appellant to home confinement with various conditions. See Order Setting Conditions for High Intensity Supervision Program, United States v. Manafort, No. 17-cr-201 (D.D.C. Oct. 30, 2017), ECF No. 9. Among the conditions, the release order stated that Appellant was “not to commit any criminal offense” while on release, and “a rearrest for any offense based upon probable cause may be grounds for revoking [] release.” Id. at 2. This condition - to not commit any further crimes was imposed yet again when his release conditions were modified a few weeks later. See Order, United States v. Manafort, No. 17-cr-201 (D.D.C. Dec. 15, 2017), ECF No. 95.

As discussed below, the alleged violation of this release condition is the core of this appeal.

In the meantime, the District Court issued a “gag order,” which commanded all parties involved in the case “to refrain from making further statements to the media or in public settings that are ‘substantially likely to have a materially prejudicial effect on this case,‘” without objection. Order (“Gag Order“) at 1, United States v. Manafort, No. 17-cr-201 (D.D.C. Nov. 8, 2017), ECF No. 38 (quoting Gentile v. State Bar of Nev., 501 U.S. 1030, 1076 (1991)). In a filing relating to conditions of release, the Government proffered that Appellant had violated the Gag Order by “ghostwriting an editorial . . . regarding his political work for Ukraine.” See Gov.‘s Opp‘n to Def.‘s Mot. to Modify Conditions of Release at 2, United States v. Manafort, No. 17-cr-201 (D.D.C. Dec. 4, 2017), ECF No. 73. On December 5, the District Court ordered Appellant to show cause why he had not violated the Gag Order. See Minute Order, United States v. Manafort, No. 17-cr-201 (D.D.C. Dec. 5, 2017). Appellant responded that he merely had “edited” the op-ed “to ensure its accuracy” and that the Ukrainian editorial was unlikely to reach an American audience, so it did not pose a substantial likelihood of a materially prejudicial effect. See Def.‘s Reply Mem. to Gov.‘s Opp‘n to Mot. to Modify Conditions of Release and Resp. to Minute Order, United States v. Manafort, No. 17-cr-201 (D.D.C. Dec. 7, 2017), ECF No. 79. The District Court discharged the show-cause order, but noted that it would “view similar conduct in the future to be an effort to circumvent and evade the requirements” of the Gag Order. Status Conference Tr. at 12, United States v. Manafort, No. 17-cr-201 (D.D.C. Dec. 11, 2017), ECF No. 112. The District Court also stated that it would consider the issue of the editorial “in connection with [its] consideration of the bond issue.” Id.

The investigation of Appellant apparently continued, because the Special Counsel filed a superseding indictment against Appellant on February 23, 2018. See Superseding Indictment, United States v. Manafort, No. 17-cr-201 (D.D.C. Feb. 23, 2018), ECF No. 202. The February 23 Superseding Indictment included new details about how Appellant allegedly acted as an unregistered agent of a foreign principal and related offenses. Specifically, the indictment alleged that Appellant and his associates retained former senior European politicians, which they referred to as the “Hapsburg group,” to advocate on behalf of certain Ukrainian principals by conducting “outreach to United States politicians and press” and by “lobbying in the United States.” Id. ¶¶ 29-30.

The indictment in our District Court is not the only federal criminal case pending against Appellant. On February 22, 2018, the Special Counsel filed a separate superseding indictment against Appellant in the United States District Court for the Eastern District of Virginia, alleging tax evasion, failure to report foreign assets, and bank fraud. United States v. Manafort, No. 18-cr-83 (E.D. Va. Feb. 22, 2018), ECF No. 9. (Apparently venue for these charges does not lie in the District of Columbia, and Appellant declined to waive venue. See Appellee‘s Br. 6 n.4.) The District Court for the Eastern District of Virginia entered a pretrial release order on March 9, 2018, which contained a condition not found in the District of Columbia release orders, namely that Appellant “must avoid all contact, directly or indirectly, with any person who is a victim or witness in the investigation or prosecution of the defendant.” Order (“EDVA Stay-Away Order“) at 3, United States v. Manafort, No. 18-cr-83 (E.D. Va. Mar. 9, 2018), ECF No. 25.

This brings us to the events directly underlying the instant appeal. On June 4, 2018, the Government moved to revoke Appellant‘s release and remand him into custody pending trial pursuant to 18 U.S.C. § 3148. See Gov.‘s Mot. to Revoke or Revise Def.‘s Current Order of Pretrial Release, United States v. Manafort, No. 17-cr-201 (D.D.C. June 4, 2018), ECF No. 315. The Government argued that Appellant violated the terms of his release “by attempting to tamper with potential witnesses” in violation of 18 U.S.C. § 1512(b)(1). Id. at 1. The core allegations of the Government‘s motion related to “[t]wo individuals who were principals of a public-relations company [that] acted as intermediaries between Manafort, [an associate then identified as Person A], co-defendant Richard Gates, and the Hapsburg group.” Id. at 4. The Government contended that Appellant “repeatedly contacted” these two witnesses - Persons D1 and D2 - “in an effort to secure materially false testimony concerning the activities of the Hapsburg group.” Id. at 5. As detailed in FBI Agent Brock Domin‘s declaration, filed with the Government‘s revocation motion, the Government alleges that Appellant attempted to contact D1 and D2 on multiple occasions from February 24, 2018, to April 4, 2018. Decl. in Supp. of Gov‘s Mot. to Revoke or Revise Def.‘s Current Order of Pretrial Release (“Domin Decl.“), United States v. Manafort, No. 17-cr-201 (D.D.C. June 4, 2018), ECF No. 315-2. These attempts included a call to D1 in which “Manafort stated that he wanted to give Person D1 a heads-up about Hapsburg,” a message Appellant sent to D1 via an encrypted application transmitting a link to a news story about the February 23 Superseding Indictment with a note that “[w]e should talk. I have made clear that they worked in Europe,” and text messages from Person A to D1 and D2 conveying that “P want[ed] to give [D1] a quick summary that he says to everybody (which is true) that our friends never lobbied in the US, and the purpose of the program was EU.” Domin Decl. ¶¶ 14-17. Other messages conveyed Appellant‘s interest in contacting D1 and D2. Id. ¶ 18.

Significantly, both D1 and D2 told Agent Domin that the Hapsburg group lobbied in the United States, not just in Europe. Id. ¶¶ 19-20. Thus, “D1 understood [Appellant‘s] messages to be an effort to ‘suborn perjury.‘” Id. ¶ 19. Similarly, D2 understood Appellant‘s outreach as an effort to get D1 to tell the members of the Hapsburg group that if they “were contacted by anyone, they should say that their lobbying work was exclusively in Europe,” id. ¶ 20, even though that was not true.

A few days after filing the motion to revoke Appellant‘s release, the Special Counsel filed a second Superseding Indictment. Superseding Indictment, United States v. Manafort, No. 17-cr-201 (D.D.C. June 8, 2018), ECF No. 318. The June 8 Superseding Indictment named Appellant‘s associate Konstantin Kilimnik as Person A, who coordinated with D1 regarding lobbying for Ukraine, and it charged Appellant and Kilimnik with obstruction of justice and conspiracy to obstruct justice in violation of 18 U.S.C. § 1512(b)(1) (“witness tampering“). Id.

Appellant opposed the Government‘s efforts to detain him. He argued that the accusations of witness tampering rested “on the thinnest of evidence” because the alleged communications did not “reflect an intent to corruptly influence a trial witness.” Def.‘s Opp‘n to Gov.‘s Mot. to Revoke or Revise the Current Order of Pretrial Release at 1-2, United States v. Manafort, No. 17-cr-201 (D.D.C. June 8, 2018), ECF No. 319. Rather, Appellant insisted that the few substantive communications were “entirely consistent with Mr. Manafort‘s stated position and repeated assertion of his innocence.” Id. at 4. Appellant also argued that he was not precluded from communicating with anyone, including possible witnesses, by the District of Columbia release orders, and he was unaware of which individuals would be potential witnesses in any event. Id. at 7-8.

After convening a hearing and allowing extensive argument by both sides, the District Court granted the Government‘s motion and ordered Appellant detained. See June 15, 2018 Hr‘g Tr. (“Hr‘g Tr.“), United States v. Manafort, No. 17-cr-201 (D.D.C. June 15, 2018), ECF No. 329. The District Court stated its findings and conclusions on the record, and it supplemented the oral findings by written memorandum opinion later that day, incorporating the reasons stated at the hearing. See Order of Detention (“Mem. Op.“) at 1, United States v. Manafort, No. 17-cr-201 (D.D.C. June 15, 2018), ECF No. 328.

First, the District Court concluded that the Grand Jury‘s indictment of Appellant on witness-tampering charges gave rise to a finding under 18 U.S.C. § 3148(b)(1)(A) of “probable cause to believe that the person has committed a Federal, State, or local crime while on release.” See Hr‘g Tr. at 19; Mem. Op. at 11. The court noted that the “commission of a crime” was the only potential violation of a condition of release that “[had] been put before [the District Court],” Hr‘g Tr. at 19, and so rested the § 3148(b)(1) finding solely on the commission of a subsequent offense while on release. The District Court declined to “get into any assessment of the strength or the deficiencies of the evidence that‘s been presented” because the Grand Jury indictment was conclusive as to probable cause. Id. at 18.

The finding of probable cause that Appellant committed an offense while on release gave rise to a “rebuttable presumption” “that no condition or combination of conditions will assure that the person will not pose a danger to the safety of any other person or the community.” Id. at 19. Because Appellant “did supply some exhibits” regarding the witness-tampering charges, the District Court found, “given the low threshold here, that [Manafort] ha[d] come forward with some evidence to rebut the presumption.” Id. at 23; Mem. Op. at 15. However, the presumption remained a relevant factor in the District Court‘s assessment of the danger Appellant may pose to the safety of the community. Hr‘g Tr. at 22; Mem. Op. at 15. The District Court also discussed the possibility that the witness contacts at issue violated the EDVA Stay-Away Order, and it explained its concerns regarding Appellant‘s November 2017 activity in editing the Ukrainian editorial while restricted by the Gag Order, which it considered “skating close to the line.” Hr‘g Tr. at 49-51.

Weighing this information, the District Court made two findings under 18 U.S.C. § 3148(b)(2). First, the District Court concluded under § 3148(b)(2)(A) that no combination of conditions could assure that Appellant would not pose a danger to the safety of the community if he remained on release.¹ The court reasoned that the witness-tampering charges indicate that Appellant poses a danger to the safety of the community through “harm to the administration of justice; harm to the integrity of the courts.” Mem. Op. at 17. While these risks are “abstract,” the court explained, the charges represent “a corrupt attempt to undermine the integrity and truth of the fact-finding process upon which our system of justice depends.” Id. Moreover, the potential harm relates to the “danger that the defendant will commit another crime of any nature while he is on release.” Id. Weighing “the nature of allegations in the superseding indictment and the evidence supplied in support of the government‘s motion . . . and the clear impact the statements had on the recipient, who reported them to the prosecution as an attempt to suborn perjury,” with the presumption arising from the Grand Jury‘s finding of probable cause for the witness-tampering charges, the court concluded “that there are no conditions that would assure that the defendant will comply with the most fundamental condition of release under the Bail Reform Act: that he not commit a Federal, State, or Local crime during the period of release.” Id. at 19.

Second, the District Court held that Appellant could not “be trusted to comply with the Court‘s directives” with respect to any conditions of release. See Hr‘g Tr. at 51; Mem. Op. at 17, 19. The District Court reasoned that “[i]t would be entirely impractical and ineffective to demand the surrender of [Manafort‘s] cell phone or to disconnect his internet service at home,” due to the difficulty of monitoring to ensure compliance. Mem. Op. at 17. The District Court further explained that it was “very troubled” that some of the alleged contacts with witnesses were made after the EDVA Stay-Away Order was in place, noting that it would “consider the defendant‘s adherence to that Court‘s admonitions in determining whether it can place its trust in the defendant.” Id. at 18. Finally, the court considered Appellant‘s request for “the issuance of an order [listing witnesses] that would absolutely be clear enough for him to follow.” Id. In the District Court‘s view, such an order would be unworkable: “there is a real risk that the defendant will interpret any list naming certain individuals as license to contact any other individuals involved in the investigation. The Court cannot draft an order that is specific enough to cover every possible future violation of the United States Code, and it should not have to.” Id. at 18-19. On this rationale, the District Court held that “it could not find, as the statute requires that it must, that defendant Manafort would abide by [any additional] conditions” of release. Id. at 19.

Having made the required findings under 18 U.S.C. § 3148(b), the District Court ordered Appellant detained.

Manafort appealed the District Court‘s detention order ten days later. He sought release pending his appeal under Federal Rule of Appellate Procedure 9(a)(3), which this Court denied. We now address Appellant‘s merits appeal.

II.

The District Court relied solely on probable cause of Appellant‘s commission of a crime to find that Appellant was subject to revocation of his release under § 3148(b)(1)(A). That finding is not contested here. The District Court then made each of the two § 3148(b)(2) findings, that the proposed package of release conditions would not assure the safety of the community under § 3148(b)(2)(A), and that Appellant is unlikely to abide by any conditions of release under § 3148(b)(2)(B). Appellant challenges each of these findings. Either finding provides an independent basis for detention, so upholding either finding is sufficient to uphold the District Court‘s detention order.

Appellant asserts that both findings are reviewed for clear error, Appellant‘s Br. 9-10, and the Government does not disagree, Appellee‘s Br. 12-13, 19. We employ the clear error standard of review because both parties ask us to do so, though we note that the standard of review for the determination that a defendant is unlikely to abide by any conditions of release remains an open question in this Court.² Because we find no clear error in the District Court‘s finding that Appellant is unlikely to abide by any conditions of release, we affirm and therefore do not reach the challenge to the dangerousness finding.

Appellant attacks three underlying factual bases upon which the District Court found that Appellant would be unlikely to abide by any conditions the District Court might craft: (1) the June 8 Superseding Indictment against Appellant for witness tampering; (2) Appellant‘s potential violation of the Stay-Away Order issued in his separate case pending in the Eastern District of Virginia; and (3) Appellant‘s actions in relation to the Gag Order issued by the District Court in the DDC case. Appellant‘s Br. 18, 19-20, 20-22.

We agree with Appellant that the District Court‘s implicit finding that his communications violated the EDVA Stay-Away Order is problematic. At the detention hearing, the District Court viewed the EDVA Stay-Away Order as sufficiently broad to cover any potential witnesses in the EDVA or DDC cases, Hr‘g Tr. at 50, and the written detention ruling, after characterizing the EDVA Stay-Away Order as “clear and unambiguous,” suggested that Appellant had not adhered to that order, Mem. Op. at 18. For several reasons, we disagree.

First, the statutory scheme supports a narrow reading of stay-away orders. Congress specified that a release order “include a written statement that sets forth all the conditions to which the release is subject, in a manner sufficiently clear and specific to serve as a guide for the person‘s conduct.” 18 U.S.C. § 3142(h)(1) (emphasis added). The statutory requirement for clarity accords with the familiar rule that the court should not punish someone for violating an order if the terms of that order are unclear. Cf. Armstrong v. Exec. Office of the President, Office of Admin., 1 F.3d 1274, 1289 (D.C. Cir. 1993) (reversing a civil contempt finding because of ambiguity in the order). The language in the EDVA Stay-Away Order is at the very least ambiguous with respect to whether the prohibition on contact with “any person who is a victim or witness in the investigation or prosecution” prohibits contact with potential witnesses to conduct underlying charges brought in a separate case in another jurisdiction. And any ambiguity in the EDVA Stay-Away Order should be construed against the Government, not Appellant.

Second, the statutory provision authorizing stay-away orders focuses on the specific offense pending before that court at that moment. Section 3142 provides, as a potential condition of release, a requirement that the defendant “avoid all contact with an alleged victim of the crime and with a potential witness who may testify concerning the offense.” 18 U.S.C. § 3142(c)(1)(B)(v) (emphases added). The use of the definite article “the” before “crime” and “offense” suggests a narrow reading of pretrial release orders, confined to witnesses to the alleged conduct charged in the particular case for which the defendant has sought pretrial release. That same definite article is incorporated into the EDVA release order. EDVA Stay-Away Order at 3 (Appellant “must avoid all contact, directly or indirectly, with any person who is a victim or witness in the investigation or prosecution of the defendant” (emphasis added)). Thus, it would be natural for anyone, including Appellant, to believe that the no-contact provision is applicable only to the particular charges brought in the EDVA case. While we agree that the Stay-Away Order could plausibly be read to cover witnesses in the District of Columbia case, such a reading is not clear and unambiguous.

Third, we find it significant that the Government never contended below that Appellant violated the EDVA Stay-Away Order. If the no-contact provisions of the order clearly applied to Appellant‘s conduct, it seems quite curious that the Government never said so below. (The Government attempts to do so now, see Appellee‘s Br. 20, but because it did not raise this argument before the District Court, it is forfeited.)

The context of the hearing indicates that the District Court concluded that the EDVA Stay-Away Order covered witnesses in the District of Columbia case. See Hr‘g Tr. at 31 (asking Appellant if it is his “position that he could call witnesses in this [District of Columbia] case, but not witnesses in that [EDVA] case?“). Further, the District Court construed D1 and D2 to be witnesses in the District of Columbia case, because all of the discussion centered around Appellant‘s attempts to discuss their testimony about the then-new allegations related to the unregistered-foreign-agent charges in the February 23 Superseding Indictment in the District of Columbia. Thus, the District Court found that Appellant violated the EDVA Stay-Away Order by contacting witnesses in the District of Columbia case, not witnesses in the EDVA case. This finding was error.

But what impact does this error have on the ultimate finding that Appellant is unlikely to abide by conditions of release in the future? When reviewing under the clear error standard, “we do not weigh each piece of evidence in isolation, but consider all of the evidence taken as a whole.” Barhoumi v. Obama, 609 F.3d 416, 424 (D.C. Cir. 2010) (quoting Awad v. Obama, 608 F.3d 1, 7 (D.C. Cir. 2010)).

Despite the District Court‘s finding that Appellant violated the terms of the EDVA Stay-Away Order, that error does not undermine the District Court‘s ultimate conclusion. Given the District Court‘s consideration of all of the information before it and its other subsidiary findings, we are not “left with the definite and firm conviction that a mistake has been committed” in the District Court‘s finding under § 3148(b)(2)(B), in light of “the entire evidence.” See Barhoumi, 609 F.3d at 423 (quoting United States v. U.S. Gypsum Co., 333 U.S. 364, 395 (1948)).

We find no clear error in the District Court‘s ultimate finding that “there are no conditions that would assure that the defendant will comply with the most fundamental condition of release under the Bail Reform Act: that he not commit a Federal, State, or local crime during the period of release.” Mem. Op. at 19. This finding was based on Appellant‘s course of conduct throughout the “six months” he had remained released in the District of Columbia case and Appellant‘s “abuse[] [of] the trust placed in” him by the District Court during that time. Hr‘g Tr. at 51. That course of conduct included Appellant‘s decision to push the envelope by contributing to an op-ed in a foreign newspaper while under the Gag Order and his repeated communications with potential witnesses, personally and through an intermediary. “[A]ll of this” conduct, which the District Court chronicled in detail during the hearing, “affect[ed] [the District Court‘s] judgment about whether [Appellant] can be trusted to comply with the Court‘s directives.” Id. The District Court did not clearly err by finding Appellant was unlikely to do so.

The conduct that loomed largest - in both the briefing on the revocation motion and in the District Court‘s findings - was the evidence suggesting Appellant had committed a crime while on release. The District Court found that if Appellant had committed a new crime even while ordered not to commit a new crime, it was likely to happen again. The District Court observed that there was no way to prevent Appellant from accessing devices that would enable him to contact witnesses while released, Mem. Op. at 17, and thus no way to ensure that further witness tampering would not occur in the future unless he were detained. After all, the communications with D1 and D2 were only discovered by the Government because the witnesses reported them; otherwise, they likely would have gone undetected. The District Court‘s core conclusion was that even if it entered an order commanding Appellant not to commit further crimes, and even if the order listed every offense in the U.S. Code, it would not be effective because Appellant would attempt to circumvent it. Id. at 18-19. It was also not lost on the District Court that Appellant had been warned about “skating close to the line” with respect to the potential violation of the Gag Order, Hr‘g Tr. at 50, and yet Appellant failed to heed those warnings and went right past the line with the alleged witness tampering.

While it is true that the District Court also considered the communications to be a violation of the EDVA Stay-Away Order, we do not find clear error after reviewing the entirety of the District Court record. The ultimate § 3148(b)(2)(B) finding was that Appellant was unlikely to abide by any conditions the District Court might impose, including the “most fundamental condition of release . . . that he not commit [additional] crime[s] during the period of release.” Mem. Op. at 19. The District Court‘s treatment of the EDVA Stay-Away Order was merely part of the icing; the cake had already been baked.

We find Appellant‘s remaining arguments to be without merit. Therefore the District Court‘s detention order is

Affirmed.

Notes

The District Court did not state by what standard of proof it made this finding, and we can resolve this appeal without delving into that issue. Cf. (detention based on communal safety reviewed to determine if supported by clear and convincing evidence, while detention based on risk of flight reviewed to determine if supported by a preponderance of the evidence). Nonetheless, district courts would be well advised to state the standard of proof being employed on the record in future circumstances, just in case. Cf. (per curiam) (remanding for further explication of reasoning where explanation in detention order was insufficient).

Although we have previously characterized a finding of dangerousness in a detention determination as a finding of fact to be reviewed for clear error, (per curiam) (citing ), we have never addressed detention based upon a finding that the defendant was unlikely to abide by conditions of release. Other circuits have taken varying approaches to review of detention orders. Some treat the ultimate determinations of dangerousness, risk of flight, or likelihood of abiding by conditions of release as factual findings to which a clear error standard of review applies. See ; (en banc); . Others have applied what has been described as “independent review” with some deference to the district court. ; . Still other courts defer to the district-court factual findings, but treat “conclusions based on such factual findings,” including the necessity of detention, as mixed questions of fact and law, reviewed de novo. See (citing ); accord (reviewing “the ultimate question whether detention is warranted” de novo). We leave resolution of these thorny questions for another day when they are fully presented and briefed.

Read the detailed case summary