United States v. Treacy

639 F.3d 32, 39 Media L. Rep. (BNA) 1897, 2011 U.S. App. LEXIS 4623, 2011 WL 799781

• Court: Court of Appeals for the Second Circuit
• Decided: March 9, 2011
• Docket: Docket 09-3939-cr
• Status: Published

Opinion

HALL, Circuit Judge:

Defendant-appellant James J. Treacy is a former Chief Operating Officer and President of Monster Worldwide, Inc. (“Monster”) — also formerly known as TMP Worldwide, Inc. — which operates the job-hunting website Monster.com. Treacy, who left his position as an officer in 2002 and left Monster’s board in 2003, was one of several Monster officials implicated in a long-term conspiracy to backdate stock options at the company and obtain favorable strike prices for the officials and others while creating the false appearance that the options had been granted at fair market value.

Treacy was indicted in August 2008 on: (1) one count of conspiracy to commit securities fraud, file false reports with the Securities and Exchange Commission (“SEC”), make false statements to auditors, and falsify books and records, in violation of 18 U.S.C. § 371; and (2) one count of substantive securities fraud, in violation of 15 U.S.C. §§ 78j(b), 78ff, 17 C.F.R. § 240.10b-5, and 18 U.S.C. § 2. A jury found Treacy guilty on both counts, and in September 2009 the district court (Rakoff, J.) sentenced Treacy principally to 24 months’ imprisonment and ordered him to pay restitution and forfeiture in the amount of $6,332,995.

On appeal, Treacy argues that his conviction must be reversed because: (1) the district court denied him his rights under the Confrontation Clause when it strictly limited his cross-examination of a Wall Street Journal reporter because of the journalist’s privilege; and (2) the district court abused its discretion in conducting voir dire when it declined to use a questionnaire agreed to by the parties and refused to question prospective jurors about their views on corporate America generally. Treacy also argues that the district court committed clear error in determining the forfeiture amount.

For the reasons that follow, we reject Treacy’s challenges to his conviction, but we agree that the district court erred in calculating forfeiture and that the case must be remanded for entry of a revised order of forfeiture.

BACKGROUND

I. Voir dire

Before trial, with the Government’s consent, Treacy provided the district court with a list of 77 questions that he proposed be given to the prospective jurors in written form. The majority of the questions on the 33-page form were general biographical inquiries that might be made in any trial. A number, however, pertained specifically to the jurors’ experiences with and views of the financial sector. Representative examples included:

• “The defendant in this case is a former President and Chief Operating Officer of a public company. Have you or has any family member or close friend ever worked as a senior-level corporate executive or as a director or officer of a public company?” Proposed Juror Questionnaire at 10.

• “Do you believe that you can be fair and impartial in a case in which a senior business executive who was paid large amounts of compensation is charged with committing a crime?” Id.

• “Do you personally know anyone who has been charged with violations of the securities laws?” Id. at 13.

• “Would the fact that the charges in the case involve the securities industry or fraud alleged to have been committed in connection with the securities industry affect your ability to be fair and impartial in this case?” Id.

• “Is there anything about what you have heard, read, seen or experienced with regard to the current economic crisis that you believe would affect your ability to be a fair and impartial juror in a case involving a former executive of a public company?” Id. at 14.

The district court refused to give the jury a written questionnaire, stating: “I never give them, ever. I did it once, and it was the biggest mistake I ever made.” The judge explained:

The[re are] many problems with questionnaires, but the single biggest problem[ ] is that there is no one in the world who can draft a question that will not have ambiguities that will be, as I learned the one time [I] tried it, that will be picked up on by various prospective jurors. And thus during the voir dire, a huge amount of time will be spent explaining to a juror why he or she misunderstood the question in the questionnaire or finding out that ... the way he interpreted it was not the way the lawyers interpreted it.

Tr. of Jan. 30, 2009, Hrg. at 12-13.

Immediately before the first venire was brought into the courtroom, counsel for Treacy stated that the two questions he considered most important were: (1) “the jurors’ knowledge of ... corporate America generally and options in particular”; and (2) “whether or not a particular juror has suffered a financial hit as a result of recent events and carries potential prejudice against people associated with Wall Street as a result.” Voir Dire Tr. at 9. The district court agreed to ask jurors about their knowledge of options, but deemed a broader inquiry to be “too remote to warrant inquiring in the broad-brush way you’ve phrased it.” Id. at 10.

The district court asked the prospective jurors, inter alia: (1) whether they could be fair and impartial in a securities or stock fraud case involving a former chief operating officer and president of Monster; (2) whether they or any of their family members or close friends had ever received any stock options; (3) whether they owned stock in or were associated with, or had a close friend or relative who worked for, Monster; and (4) whether they had been exposed to any pretrial publicity about the case. Where prospective jurors gave affirmative answers to any of these questions, the district court probed further for possible bias. When one prospective juror mentioned that a “very close friend” had been “burned by the Bernie Madoff’ case, the district court stated: “This has absolutely nothing to do with the Bernie Madoff case or anything like it.” Id. at 44. The district court admonished the jury: “[WJe’ve all read about cases, high visibility economic cases; a number of them have come up in this very court. This has nothing to do with those cases. It is a totally separate, different kind of case, and you shouldn’t even think for one second that it has anything to do with those cases whatsoever.” Id. at 44-45.

II. Trial

i. Evidence of the backdating scheme

Monster’s vice president for human resources, Erin Barriere, testified how stock options typically work at Monster. She explained that a stock option usually gives the holder the right to buy a share at a set price, known as the “strike” price or “exercise” price, and that this price is normally the fair market value of the stock on the day the option is granted. Trial Tr. at 88-91, 126. The typical vesting period for an option at Monster was four years, with 25 percent of the granted shares vesting after each one-year interval.

Cooperating witness Myron Olesnyekyj testified that he worked at Monster from 1994 until 2006, when he was fired because of his participation in the backdating scandal. Olesnyekyj testified that he helped put together documentation relating to the backdating and hid information from directors, auditors, and the public. He testified to how the scheme worked:

A. By backdating I mean that we would select a price with the benefit of hindsight. We would look back for some period of time to select a favorable low price, and instead of having the documents reflect the actual market prices on the date of selection of that price by the compensation committee, I would create fictitious documents, suggesting that price was selected on the date of the low price.

Q. Did the date you or others selected have anything to do with the date the grants were actually approved by the compensation committee?

A. It did not.

Q. What did Monster tell its auditors and the public about how the exercise price was being chosen for option grants during the period 1997 through 2006?

A. Monster told its auditors and the public that the price reflected the price on the actual date of approval by the compensation committee, and that price reflected the fair market value on the date of approval by the compensation committee.

Q. Were those representations accurate?

A. They were not....

Q. What was the purpose of the backdating of option grants?

A. The purpose of the backdating of option grants was to provide in-the-money options at low prices without taking the required expense for those options---- That served the purpose of making the company’s earnings and profits look a lot better than they were. Trial Tr. at 315-16.

During the years in question, Monster filed public documents with the SEC claiming that it followed Accounting Prin ciples Board Opinion No. 25 (Oct. 1972), titled “Accounting for Stock Issued to Employees” (“APB 25”). As effective during the relevant years, APB 25 required a company to record any expense or charge on its earnings for the value of an employee stock option on its “measurement date,” defined as the date on which the officials at the company who were authorized to confer stock options approved and issued a known and specified number of options at a known and specified price. Because of this requirement, whenever a company issued any options “in the money” (meaning, any options with a strike price lower than the fair market value of the stock on the measurement date), it was required to report a charge on its earnings in the form of a compensation expense, apportioned over the vesting period.

Under Monster’s official stock option policy, no members of management — Treacy included — had authority to grant options, whether acting individually or collectively. Rather, exclusive authority to grant options was vested in a Compensation Committee comprising independent directors, all of whom had to sign a unanimous written consent form (“UWC”) to confer an option, and every UWC had to specify an “as of’ date, meaning the date on which the option was deemed to have been granted. The Government introduced evidence at trial to show that Treacy and his co-conspirators granted options to themselves and others with strike prices often far below the fair market value of the stock on the day of the grant, and did so by falsifying UWCs with “as of’ dates chosen in hindsight based on low closing prices and not based on the dates of any actual action taken by the Compensation Committee. By backdating the UWCs, the conspirators could avoid the requirements of APB 25 and compensate themselves without reducing Monster’s reported earnings.

Copies of e-mails introduced at trial showed that Treacy often voted on when to set “as of’ dates and vesting dates, and approved backdating proposals by other conspirators. The prosecution also introduced documents that Treacy had signed, including SEC filings and management representation letters that contained false statements. Treacy’s own profit was significant — he received more than one million options in eight different grants, six of which were backdated.

ii. Testimony of Wall Street Journal reporter Charles Forelle

The prosecution also introduced a June 12, 2006, story from the Wall Street Journal (“WSJ”), co-written by reporter Charles Forelle and titled “Monster Worldwide Gave Officials Options Ahead of Share Run Ups.” The story stated that the WSJ had “put[ ] the odds at about one in nine million that a pattern of grants as favorable or more favorable than Mr. Treacy’s would have occurred if the dates were selected randomly, without regard to stock price.” The story contained the following passages, among others:

Mr. Treacy, who has since left Monster, said that, like any other employee, he had no involvement in the options-granting process, which he said was ‘all in the purview’ of the board’s compensation committee and Andrew McKelvey, Monster’s founder and chief executive. He said he believes the dates were ‘the days that the comp committee and Andy granted’ the options....

Mr. Treacy said that he didn’t notice the favorable strike price at the time. “I was busy working, and there was a lot to do and a lot of moving parts and a family to get home to,” he said.

He also said that not all of his grants were unusual. A 2001 grant came be fore a fall in share price, and another grant was dated on the same day the company publicly announced his promotion to president. He left the company in 2002, but remained on the board until 2003.

Mr. Treacy never exercised any of his options while he was an executive. As part of his separation agreement, the company allowed him to keep them after he left. He cashed out some for the first time in December [2005], he said. Asked how much he made on the transaction, Mr. Treacy said he doesn’t have to disclose the sum. “I did OK,” he said. “But I waited a long time.”

Charles Forelle and Mark Maremont, Monster Worldwide Gave Officials Options Ahead of Share Run-Ups, Wall St. J., June 12, 2006, at A1.

At trial, the Government subpoenaed Forelle to testify to the accuracy of the statements attributed to Treacy in his story. Forelle sought to quash the subpoena, invoking the journalist’s privilege and claiming that his testimony would be irrelevant, cumulative and available from other sources. The Government argued that Treacy’s statements to Forelle were made in furtherance of the conspiracy, showed his consciousness of guilt, and demonstrated his knowledge of the stock options process at Monster. The Government also argued that Forelle’s testimony was the only means by which the statements could be authenticated.

At oral argument on Forelle’s motion to quash, the Government asked to be permitted to ask questions primarily to confirm the substance and accuracy of Treacy’s statements in the story. Treacy argued that he should be able to cross-examine Forelle regarding: (1) the specific questions that led to the statements attributed to Treacy; (2) statements attributed to McKelvey; and (3) Forelle’s other reporting on backdating schemes, including those at other companies. Treacy argued that he wished to show that his statements to Forelle were about his own stock options, not about options-granting generally.

The district court denied Forelle’s motion to quash but “tightly limit[ed]” both direct and cross-examination of the reporter. Dist. Ct. Mem. Ord. dated Mar. 23, 2009, at 2. The court rejected Forelle’s contention that the evidence was cumulative. Applying our decision in Gonzales v. Nat’l Broadcasting Co., 194 F.3d 29, 36 (2d Cir.1999), the district court noted that the testimony sought was nonconfidential and concluded that the statements were “relevant, at a minimum, as statements made in furtherance of the alleged conspiracy and false exculpatory statements evidencing consciousness of guilt, and may also prove admissible for other purposes.” Id. at 4. But the court limited direct examination to “four essential inquiries: (1) In or around June 2006, was Forelle a reporter for the Wall Street Journal? (2) Did Forelle interview [Treacy] during that time? (3) Did [Treacy] make statements to Forelle that Forelle subsequently reported ... ? [and] (4) With respect to each of those statements, what specifically did Forelle say to [Treacy] and what specifically did [Treacy] say in response?” Id. The district court, noting that Treacy had not argued that the statements had not been made or misreported, but only that they were taken out of context, limited his cross-examination to asking about the questions Forelle had asked Treacy immediately before those that elicited the responses quoted in the story.

Called to testify, Forelle confirmed that Treacy had made the statements in the WSJ article. On cross-examination, Treacy’s counsel asked Forelle if “what you asked him about during that conversation was Mr. Treacy’s options,” and Forelle replied, “It’s correct that I asked him about his options.” Trial Tr. at 2240. Treacy’s counsel then asked: “And you asked him specifically — the reason you called him is because you saw his options in public filings and you wanted to ask him about it, isn’t that true?” Id. at 2241. The district judge sustained an objection to that question from the reporter’s counsel, whom the court had previously granted permission to appear and raise silent objections where appropriate. At a sidebar conference, the district court explained that, because of the need to balance press interests with the parties’ interests, Treacy could not make an open-ended attack on Forelle’s credibility. Treacy attempted to ask Forelle several times about a follow-up e-mail that Forelle sent to Monster’s public relations consultant in which he mentioned the interview with Treacy, but the district court ruled the questions improper. On redirect, the Government asked Forelle, “were your questions to Mr. Treacy limited to his own grants or was it more broadly the grants and practices at the company as a whole?” Id. at 2249. Forelle replied, “It was both.” Id. On inquiry from the bench, Forelle was unable to recall the specific questions he had asked that had elicited the statements by Treacy quoted in the story. Although Treacy was prohibited from cross-examining Forelle based on the follow-up e-mail Forelle had sent to Monster’s public relations contact, the district court did allow Treacy to introduce the e-mail itself, which contained the following passage:

Mr. Treacy says that his options’ terms were all decided by the compensation committee and Mr. KcKelvey. Is this accurate? He characterizes Mr. McKelvey’s role in option-awarding as principal — that “Andy gave” grants to senior executives. Could you expand on the role of Mr. McKelvey in determining the dates, amounts and nature of executive awards?

E-mail from Charles Forelle to Peter Duda dated June 10, 2006.

In its summation, the prosecution mentioned Forelle’s article at several different points, and argued that Treacy’s “quotes in the Wall Street Journal, after the news of the backdating broke, that like any other employee, he had no involvement in the options dating process,” were “a ‘monster’ lie” and a “bald-faced lie.” Trial Tr. at 2742, 2810. The Government directly tied this observation to its broader case that “[ijnstead of doing the right thing, Treacy lied over and over again.” Id. at 2747. Treacy’s counsel, in turn, argued in summation that the phrasing of Forelle’s follow-up e-mail (“Mr. Treacy says that his options’ terms were all decided by the compensation committee and Mr. McKelvey, is that accurate?”) showed that: (1) McKelvey was the principal force behind the backdating scheme; and (2) “the conversation with the Wall Street Journal reporter was not about Treacy’s relationship with the option program in toto but was specific to Treacy’s grants.” Trial Tr. at 2900-02.

iii. Forfeiture 1

The Government sought forfeiture under 18 U.S.C. § 981(a)(1)(C), which authorizes forfeiture of “[a]ny property ... which constitutes or is derived from proceeds traceable to” a multitude of criminal offenses, including securities fraud. 18 U.S.C. § 981(a)(1)(C). Invoking APB Opinion 25, the Government argued that the “measurement date” for determining the cost of stock options was “the first date on which are known both (1) the number of shares that an individual employee is entitled to receive and (2) the options or purchase price, if any.” The Government supplied a chart showing the measurement dates as modified by the trial evidence in the manner most favorable to Treac/s position, which calculated a profit of $6,332,995.

With respect to the option grant dated “as of’ January 6, 1997, the Government’s chart proposed a measurement date of April 10, 1997. The Government pointed to testimony by Olesnyckyj, who stated that the date and price for this option grant were not selected until at least the end of January 1997, and that the recipients of the grants were not fully determined until April 1997. The Government also presented a facsimile sent on April 10, 1997, which included an unexecuted Unanimous Written Consent (“UWC”) form of the Compensation Committee for a grant of 1.8 million options dated “as of’ January 6, 1997 (66,666 of which were granted to Treacy).

With respect to the grant with an “as of’ date of December 9, 1998, the Government’s chart proposed a measurement date of April 19,1999. In support of this measurement date, the Government relied upon a series of e-mails dated April 12, 1999, in which Treacy, Olesnyckyj, and Monster’s controller discussed the number of options to authorize under Monster’s then-new stock option plan, pursuant to which the options backdated to December 9, 1998 would be issued. On or about April 19, 1999, as evidenced by the “modify date” in the relevant computer file, Olesnyckyj created a UWC for the broad-based grant with the grant date December 9, 1998.

The district court found by a preponderance of the evidence that the Government’s chart accurately reflected the evidence adduced at trial, 2 and specifically rejected Treacy’s contention that two of the grants had not been backdated at all. The district court ordered Treacy to forfeit $6,332,995, the amount in the Government’s chart.

DISCUSSION

I. The Confrontation Clause and the journalist’s privilege

Treacy contends that the strict limitations the district court placed on his cross-examination of Forelle violated his rights under the Confrontation Clause. Specifically, he argues that: (1) prescribed questioning of Forelle was no substitute for adversarial examination; (2) the Confrontation Clause requires that the defense be able to attack credibility; and (3) because the Government argued repeatedly to the jury that Treacy was “denying any involvement at all in Monster’s stock options process” — a notion that the defense says is “preposterous” — Treacy should have been able to test Forelle’s testimony by attempting to show that Treacy was only speaking of his own options. Treacy Br. at 36-37.

We review a district court’s decision to restrict cross-examination for abuse of discretion, even where the basis for challenging those restrictions is the Confrontation Clause. See, e.g., United States v. Figueroa, 548 F.3d 222, 226-27 (2d Cir.2008). When that decision “rests on an error of law (such as the application of the wrong legal principle),” a district court necessarily abuses its discretion. Id. (quoting Zervos v. Verizon N.Y., Inc., 252 F.3d 163, 169 (2d Cir.2001)).

It is settled law in this Circuit, at least in the civil context, that a journalist possesses a qualified privilege protecting him or her from the compelled disclosure of even nonconfidential materials. See Chevron Corp. v. Berlinger, 629 F.3d 297, 306-07 (2d Cir.2011); Gonzales v. NBC, 194 F.3d 29, 35 (2d Cir.1999). This is so because

even where there [is] no issue of betrayal of a promised confidence, ... “wholesale exposure of press files to litigant scrutiny would burden the press with heavy costs of subpoena compliance, and could otherwise impair its ability to perform its duties — particularly if potential sources were deterred from speaking to the press, or insisted on remaining anonymous, because of the likelihood that they would be sucked into litigation” [and] ... unrestricted litigant access to press files would create socially wasteful incentives for press entities “to clean out files containing potentially valuable information lest they incur substantial costs” of subpoena compliance, and would risk “the symbolic harm of making journalists appear to be an investigative arm of the judicial system, the government, or private parties.”

Chevron Corp., 629 F.3d at 307 (quoting Gonzales, 194 F.3d at 35 (internal citations omitted)).

No party or amicus argues that any of the information sought from Forelle was confidential, or that he is protecting the identity of any source. Forelle, therefore, is not entitled to invoke the stronger privilege that protects confidential materials. Cf United States v. Cutler, 6 F.3d 67, 71 (2d Cir.1993) (where a party seeks confidential material from a journalist, “disclosure may be ordered only upon a clear and specific showing that the information is: [ (1) ] highly material and relevant, [ (2) ] necessary or critical to the maintenance of the claim, and [ (3) ] not obtainable from other available sources”). On the contrary, not only was Forelle not protecting any confidential material or source, he sought to withhold evidence that his source himself (Treacy) desired be disclosed. Thus, because “the protection of confidential sources is not involved, the nature of the press interest protected by the privilege is narrower” and the privilege is “more easily overcome.” Gonzales, 194 F.3d at 36. “Where a civil litigant seeks nonconfidential materials from a nonparty press entity, the litigant is entitled to the requested discovery notwithstanding a valid assertion of the journalists’ privilege if he can show that the materials at issue are of likely relevance to a significant issue in the case, and are not reasonably obtainable from other available sources.” Id.

Before we review the manner in which the district judge curtailed cross examination, we will address the argument of amicus Dow Jones & Company (“Dow Jones”), parent company of the Wall Street Journal, which urges us to hold that the district court ought to have quashed the subpoena to Forelle outright. Although appearing in support of neither party and taking no position with respect to whether Treacy’s conviction should be reversed, Dow Jones argues that the bar should be higher for overcoming the reporters’ privilege in criminal cases, and that we should apply the “clear and specific showing” standard — used in civil cases only when confidentiality is at stake — in all criminal cases.

We are not persuaded. Dow Jones argues that its proposed rule is necessary to avoid a constitutional paradox that arises when a court must balance a reporter’s First Amendment rights with a criminal defendant’s Sixth Amendment rights, but this argument depends on the assumption that the reporter’s privilege is derived from the First Amendment rather than a federal common law of privileges, a question we have previously declined to resolve absent any Congressional retrenchment of the privilege. See Gonzales, 194 F.3d at 35 n. 6. We once again decline to wade into these constitutional waters, because absent from Dow Jones’s brief is any convincing reason why the Gonzales test should not apply in criminal cases where the prosecution seeks to introduce nonconfidential materials. Dow Jones suggests that because the Department of Justice has adopted strict guidelines for itself with respect to subpoenaing reporters, see 28 C.F.R. § 50.10, this Court should adopt the same principles as binding propositions of law. Nothing in our case law suggests, however, that the journalist’s privilege is stronger in criminal cases. If anything, in fact, we have recognized that our Court once set too high a bar for overcoming the privilege in criminal cases and consciously lowered that bar. See United States v. Cutler, 6 F.3d 67, 73 (2d Cir.1993); United States v. Burke, 700 F.2d 70, 76-78 (2d Cir.1983); see also Gonzales, 194 F.3d at 34 n. 3 (observing that Cutler recognized that Burke “undervalued the needs of criminal defendants”). We now hold that, in instances where a reporter is not protecting a confidential source or confidential materials, the showing required to overcome the journalist’s privilege is the same in a criminal case as it is in a civil case— namely, the showing required by Gonzales — and that this is true whether the party seeking to overcome the privilege is the prosecution or the defense. 3

To the extent that it denied Forelle’s motion to quash, while limiting the scope of the Government’s direct examination of the reporter, the district court applied Gonzales correctly. As that court explained, false exculpatory statements can provide circumstantial evidence of guilt, and this satisfies the Gonzales “likely relevance” test. Because Treacy possessed an absolute Fifth Amendment right not to testify himself, Forelle was the only potential witness who could confirm that Treacy had made the statements quoted in the article, satisfying the requirement that the material be “not reasonably obtainable from other sources.” Finally, by restricting the questions the Government could ask on direct examination, the district court protected the privilege, tailoring the questions to the showing of relevance and necessity made by the Government.

The limitations the district court imposed on the scope of cross-examination, however, are another matter. Treacy and his supporting amicus, the National Association of Criminal Defense Lawyers, argue that once a trial court has determined that the Government has made the required showing to overcome the journalist’s privilege and compel a reporter’s direct testimony, the trial court may not, consistent with the Sixth Amendment’s Confrontation Clause, thereafter employ the privilege to restrict the defendant’s cross-examination of the reporter to a greater degree than it would restrict such cross-examination in a case where no privilege was at issue. We agree.

During Forelle’s testimony, when Treacy’s counsel attempted to challenge Forelle’s credibility by bringing up Forelle’s follow-up e-mail to Monster’s public relations representative, the district court sustained objections from Forelle’s counsel, stating: “In the balancing of various competing considerations here, I did not open it up to a gerieral attack on credibility. Anything can relate to credibility, and I limited the scope much more specifically.” Trial Tr. at 2244-45. Defense counsel argued that “[t]his relates to his writing of the article” and that “this isn’t just general impeachment here.” Id. at 2245. The judge replied: “It was always the defense position that if he recalled at all, you should be able to get into the whole interview, both to establish context and to test credibility. And the [cjourt, because it had to consider the third party’s interests here, namely, the press interests and the press privilege, said no, you can’t do that.” Id.

The Confrontation Clause guarantees a criminal defendant the right “to delve into the witness’ story to test the witness’ perceptions and memory” and “to impeach, i.e., discredit, the witness.” Davis v. Alaska, 415 U.S. 308, 316, 94 S.Ct. 1105, 39 L.Ed.2d 347 (1974). Although the scope of cross-examination is “generally within the sound discretion of the trial court,” United States v. Pedroza, 750 F.2d 187, 195 (2d Cir.1984), “the trial judge’s discretion ‘cannot be expanded to justify a curtailment which keeps from the jury relevant and important facts bearing on the trustworthiness of crucial testimony,’ ” id. at 196 (quoting Gordon v. United States, 344 U.S. 414, 423, 73 S.Ct. 369, 97 L.Ed. 447 (1953)).

Once the district court made the (correct) decision that the Government had sufficiently overcome the Gonzales privilege to compel Forelle’s testimony, it was an abuse of discretion to restrict Treacy from cross-examining Forelle — subject to ordinary rules regarding the scope of direct and relevance, see Fed.R.Evid. 611(b) — as he would any other witness. The only privilege Forelle possessed in this case was the qualified Gonzales privilege, and the question before the district court during cross-examination was the same as on direct examination, namely, whether the answers defense counsel sought were of “likely relevance to a significant issue in the case, [and] not reasonably obtainable from other available sources.” The district court committed an error of law when, instead of applying the test we set forth in Gonzales to evaluate Treac/s need for Forelle’s answers, it treated Forelle’s interest as a’ competing interest to be balanced against Treacy’s Confrontation Clause rights.

It will generally be the case that the defense can satisfy the Gonzales test on cross-examination to the extent that it seeks to challenge a reporter’s credibility about the specific content of his direct testimony. We agree, of course, with the district court’s observation that “[anything can relate to credibility,” and the district court possessed the discretion to prevent a “general attack on credibility.” But we agree with Treacy that, because the purpose of Forelle’s testimony was to confirm the accuracy of the statements in the Wall Street Journal article, attempts to test Forelle’s memory with respect to the writing of the article were not a general attack on credibility but a challenge to the specific details of Forelle’s direct testi mony. Such a challenge is of “likely relevance to a significant issue in the case,” and no person but Forelle could provide the required answers. With the Gonzales test satisfied, the district court erred in balancing the defendant’s Confrontation Clause rights against a general press interest.

It is the law in this Circuit that if a witness’s invocation of the Fifth Amendment privilege against self-incrimination prevents the defendant from cross-examining the witness with respect to a non-collateral matter, the witness’s direct testimony must be stricken. See Bagby v. Kuhlman, 932 F.2d 131, 135 (2d Cir.1991); Dunbar v. Harris, 612 F.2d 690, 693 (2d Cir.1979); United States v. Cardillo, 316 F.2d 606, 611 (2d Cir.1963). A general attack on credibility is, concededly, a collateral matter. See Cardillo, 316 F.2d at 611. But testimony should ordinarily be stricken when the invocation of the privilege against self-incrimination prevents the defendant from cross-examining the witness with respect to his credibility regarding the specific details of his direct testimony. See id. at 613. And if the Confrontation Clause requires this result when a constitutional privilege is involved, it must a fortiori require the same result when a witness invokes a privilege that may derive only from federal common law, and the weakest variation on the common law privilege, at that. In other words, if the district court had believed that Treacy could not fully exercise his Confrontation Clause rights because of Forelle’s assertion of the privilege, it ought to have granted Forelle’s motion to quash or subsequently stricken his testimony.

However, “[e]ven if we conclude that the trial judge has improperly curtailed cross-examination in violation of the defendant’s confrontation rights, we are not to reverse automatically but are to apply harmless-error analysis, in order to determine Vhether, assuming that the damaging potential of the cross-examination were fully realized,’ the error was harmless beyond a reasonable doubt.” United States v. Rosa, 11 F.3d 315, 336 (2d Cir.1993) (quoting Delaware v. Van Ars dall, 475 U.S. 673, 684, 106 S.Ct. 1431, 89 L.Ed.2d 674 (1986)). “That is, we should not reverse if we conclude that the error was ‘unimportant in relation to everything else the jury considered on the issue in question.’ ” Id. (quoting Yates v. Evatt, 500 U.S. 391, 403, 111 S.Ct. 1884, 114 L.Ed.2d 432 (1991)). The factors to be considered on harmless error review include “ ‘the importance of the witness’ testimony in the prosecution’s case, whether the testimony was cumulative, the presence or absence of evidence corroborating or contradicting the testimony of the witness on material points, the extent of cross-examination otherwise permitted, and ... the overall strength of the prosecution’s case.’ ” Cotto v. Herbert, 331 F.3d 217, 254 (2d Cir.2003) (quoting Van Ars dall, 475 U.S. at 684, 106 S.Ct. 1431).

Here, even though the prosecutor repeatedly emphasized Treacy’s statements to Forelle as evidence of his mendacity, any Confrontation Clause error was harmless beyond a reasonable doubt because the other evidence in the prosecution’s case was vastly more significant to demonstrating Treacy’s actual actions. First, although Treacy was unable to cross-examine Forelle based on the e-mail, he was able to introduce the e-mail itself. He was therefore able to make — if somewhat less forcefully than he could have if permitted to engage in effective cross-examination— the argument during summation that his comments to Forelle were taken out of context. Second, this Court has recognized that false exculpatory statements evincing consciousness of guilt are only circumstantial evidence of guilt, and weak circumstantial evidence at that. See United States v. Lorenzo, 534 F.3d 153, 161 (2d Cir.2008). Concededly, the Government, at summation, did repeatedly point to Treacy’s statements as evidence of a guilty mind, but it is highly unlikely that it would have been unable to secure a conviction in the absence of Forelle’s testimony.

In addition to the particularly inculpatory testimony from Olesnyckyj, the Government introduced significant documentary evidence showing that Treacy actively voted on and approved backdating of documents. Particularly devastating is an exchange of e-mails between Olesnyckyj and Treacy in April 1999, in which Olesnyckyj, inquiring about how many options could plausibly be granted beyond the 3 million already discussed, wrote: “I think we should authorize the highest number we can get away with — what number do you think will fly with the investment community?” E-mail from Myron Olesnyckyj to Jim Treacy (Apr. 12, 1999 11:20 a.m.). Treacy replied, “Go to 15, up 12!”, E-mail from Jim Treacy to Myron Olesnyckyj (Apr. 12, 1999 2:27 p.m.), and four days later added: “Spoke to Andy last evening, and Greg this morning, and all decided to go even higher. The new plan should be $15 million. So, together with the old plan, it would total $18 million. Please confirm this gets handled correctly for the upcoming meeting,” E-mail from Jim Treacy to Myron Olesnyckyj (Apr. 16,1999 5:51 p.m.).

In short, even if Treacy had been able to persuade the jury that Forelle’s memory of their conversation was hazy, and that Treacy had only been discussing his own options, not a general backdating scheme, the other evidence at trial demonstrated that Treacy was, in fact, involved broadly in the backdating of options at Monster, and we are confident that the jury would not have been persuaded otherwise by an ambiguous newspaper article. We find the district court’s error harmless beyond a reasonable doubt.

II. Voir dire

In conducting voir dire, a district judge possesses “ample discretion.” United States v. Quinones, 511 F.3d 289, 299 (2d Cir.2007) (quoting Rosales-Lopez v. United States, 451 U.S. 182, 189, 101 S.Ct. 1629, 68 L.Ed.2d 22 (1981)). We will not interfere with the manner in which the district court has chosen to conduct voir dire unless an abuse of discretion is “clear.” Id. (quoting United States v. Salameh, 152 F.3d 88, 121 (2d Cir.1998)). Although we have approved the use of questionnaires as one of many tools available for voir dire, we have expressly declined to “hold that district judges are ever obligated to make use of this procedure in selecting juries.” Id. at 300 n. 8; see also Salameh, 152 F.3d at 120-21 (finding that the district court did not abuse its discretion in declining to use a prospective juror questionnaire).

We observed in 2002 that we “appear[ ] never to have reversed a conviction for the failure to ask a particular question on the voir dire of prospective jurors.” United States v. Lawes, 292 F.3d 123, 129 (2d Cir.2002). Research does not reveal any cases since Lawes in which that has changed. The Court in Lawes stated that

a voir dire may be so insufficient as to call for a reversal, but the record viewed as a whole must show either: (i) a voir dire so demonstrably brief and lacking in substance as to afford counsel too little information even to draw any conclusions about a potential jurors general outlook, experience, communication skills or lifestyle; (ii) a failure to inquire about, or warn against, a systemic or pervasive bias, including one that may be short-lived but existent at the time of trial, in the community that would have been cured by asking a question posed by a party; of (iii) a record viewed in its entirety suggesting a substantial possibility that a jury misunderstood its duty to weigh certain evidence fairly that would have been clarified by asking a requested voir dire question....

Id. (internal citations omitted).

Of these possible grounds, only the second is arguably applicable here. Treacy reasonably asserts that, in the spring of 2009 when he was tried, animosity toward high-ranking corporate executives was considerable, and it is at least arguable that this constituted a “systemic or pervasive bias, including one that may be short-lived but existent at the time of trial.” Id. Nevertheless, the district judge did not abuse his extensive discretion in declining to issue a questionnaire or specifically interrogate jurors about corporate malfeasance in general. Although his refusal to use questionnaires was categorical rather than based on the circumstances of this case, it cannot be termed an abuse of discretion because he offered a rational, non-arbitrary reason for his policy — that, in his experience, jurors tended to understand written questions differently from those who drafted the questions, leading to substantial difficulty in parsing their responses. Given this Court’s express disavowal in Quinones of the notion that questionnaires could ever be a required part of voir dire, see 511 F.3d at 300 n. 8, and this Court’s statement in Lawes that the methodology of voir dire is not only within the district court’s discretion but “quintessentially” so, 292 F.3d at 128, Judge Rakoffs personal policy is well within his bailiwick as a trial judge so long as he conducts adequate voir dire by some other means.

Treacy argues that even if the district court was not obliged to issue its questionnaire, his conviction should be reversed because the district court, in conducting an oral voir dire, failed to inquire broadly of the prospective jurors about their biases, if any, against corporate America generally. Treacy’s argument is unpersuasive. As Lawes implicitly recognized, a district court may find that warning a jury against an improper bias may be more effective in some cases than inquiring about that bias. See 292 F.3d at 129 (suggesting that a district court has the option either “to inquire about, or warn against, a systemic or pervasive bias” (emphasis added)). Here, the district court specifically inquired about the prospective jurors’ knowledge of, and experience with, stock options. When one juror mentioned the Bernie Madoff case, the district judge admonished the venire that this case was completely different from cases such as Madoff s. Thus, even assuming arguendo that bias against corporate executives was so “systemic or pervasive” that some warning was required, the district court made it abundantly clear to all prospective jurors that they must put this bias out of their minds. The district court may well have believed that questioning the jury about corporate bias could do more harm than good, and falsely planted in their minds the notion that this case was, in fact, tied to the economic collapse of 2008. That alone is sufficient to justify the district court’s use of a warning, rather than questioning, to guard against bias on this point, especially given the considerable range of discretion within which the court was operating.

III. Forfeiture calculation

When a forfeiture award is challenged on appeal, this Court reviews the district court’s legal conclusions de novo and its factual findings for clear error. See United States v. Sabhnani, 599 F.3d 215, 261 (2d Cir.2010). The calculation of forfeiture amounts is not an exact science. “[T]he court ‘need not establish the loss with precision but rather need only make a reasonable estimate of the loss, given the available information.’ ” United States v. Uddin, 551 F.3d 176, 180 (2d Cir.2009) (quoting United States v. Carboni, 204 F.3d 39, 46 (2d Cir.2000)). A court is permitted to use general points of reference as a starting point for calculating the losses or gains from fraudulent transactions and may make reasonable extrapolations from the evidence established by a preponderance of the evidence at the sentencing proceeding. Id. at 180.

The Government sought forfeiture under 18 U.S.C. § 981(a)(1)(C), which authorizes forfeiture of “[a]ny property ... which constitutes or is derived from proceeds traceable to” a multitude of criminal offenses including securities fraud. 18 U.S.C. § 981(a)(1)(C). “Because criminal forfeiture is viewed as part of the sentencing process, the government need prove facts supporting forfeiture only by a preponderance of the evidence.” United States v. Gaskin, 364 F.3d 438, 461 (2d Cir.2004) (internal citations omitted).

To begin with, we reject Treacy’s argument that the measurement dates for stock options at Monster were set when McKelvey made commitments to issue options, rather than when the Compensation Committee acted. To embrace this argument would be to undermine the very principles of transparent corporate governance embodied in APB 25. As Olesnyckyj testified at trial, Monster’s plan for granting stock options was ratified by shareholders and required that options be granted only by the Compensation Committee, consisting of “two non-employee or independent directors” who “weren’t employed by the company, [and did not] have significant ties to the company as defined by SEC rules.” Trial Tr. 437-38. Olesnyckyj further testified that the Compensation Committee never delegated its authority and could only act by unanimous written consent. If we were to accept Treacy’s argument that McKelvey — who was, without dispute, one of the key masterminds of the backdating scheme — possessed the actual authority to bind Monster to issue options, notwithstanding a shareholder-approved policy placing this authority in independent hands, we would not only retroactively endorse this defiance of the company’s own rules but give Treacy the financial benefit of the scheme. We conclude that APB 25 requires that the measurement date be the date on which the officials at the company actually authorized to confer stock options approved and issued a known and specified number of options at a known and specified price. With that in mind, we turn to the two specific grants that Treacy challenges.

i. The January 6, 1997, Grant

Treacy argues that because Olesnyckyj testified that the price for the grant was determined in late January, the measurement date should be late January. This factor, however, only meets the second prong of the APB standard. Olesnyckyj testified that the recipients of the grant were not determined until April. The only evidence presented that Treacy was identified as a recipient of the grant before April is the memo Treacy wrote (dated January 6), which Olesnyckyj testified he did not receive until April 1997. Thus, the District Court did not err in its conclusion that the Government proved by a preponderance of the evidence that the measurement date was April 10,1997.

ii. The December 9, 1998, Grant

The Government’s chart determined that the measurement date, April 19, 1999, was appropriate for the “as of’ December 9, 1998, grant, and the district court appears to have adopted this date as well. The Government argues that because Treacy received his 75,000 options as part of a broad-based grant of 862,000 options to multiple recipients and not as a one-off grant to him alone, the APB two-prong standard was not satisfied until all the recipients were allocated. Recipient allocations remained uncertain until at least the end of March 1999.

The Government also points to the emails dated April 12, 1999, in which Treacy and Olesnyckyj discussed the number of options to authorize under Monster’s then-new stock option plan, pursuant to which the options backdated to December 9, 1998, would be issued. See supra, p. 41-42. On or about April 19, 1999, as evidenced by the “modify date” of a computer document, Olesnyckyj created a UWC for the broad-based grant with the grant date of December 9, 1998. On this basis, the Government contends that the district court properly found the measurement date of the grant was April 19, 1999, because it was the earliest date upon which the Compensation Committee could have approved the grant with the specified recipients and number of options.

In contrast, Treacy argues that the first prong of the APB standard was met on September 11, 1998, as evidenced by a September 15, 1998, memo to Olesnyckyj in which Treacy stated, “[McKelvey] met with me on Friday, September 11, 1998[,] to advise me” that “[i]n December 1998 at annual option grant time, I will receive 75,000 TMPW share options.” According to Treacy, the second prong of the APB standard — the price of the grant — was determined on, or shortly after, January 15, 1999. At the very latest, Treacy contends that on February 9, 1999, he submitted an SEC Form 5 disclosing the quantity and the price of the December 9, 1998, option grant to him — thereby meeting the APB standard.

We cannot give any weight to Treacy’s September 15, 1998, memo because of concerns about fraud. As we have previously stated, moreover, the measurement date cannot have been set by McKelvey unilaterally. But Treacy’s second argument is persuasive. The evidence indicates that Monster issued different “rounds” of stock option grants with “as of’ dates of December 9, 1998. For example, the record shows that on February 26, 1999, Monster sent a letter to over 50 employees that they would receive a stock option grant at the strike price of $26.875 — the price of the December 9, 1998, “as of’ date stock option grant. Subsequent emails in April 1999 between Treacy and Olesnyckyj discussed a “new plan” of stock option grants with the same backdated “as of’ December 9, 1998, date and strike price of $26,875 per share. This “new plan” increased the stock options by $15 million.

As noted, the evidence shows that Treacy submitted an SEC Form 5 dated February 9, 1999, disclosing the quantity of stock options and price of the December 9, 1998, option grant. Although there were later rounds of stock option grants that used the “as of’ December 9, 1998, date, this does not change the fact that Treacy’s grant was set earlier. There is nothing in the APB standard that suggests that every recipient of the options grant with the same backdated “as of’ date must be determined to establish the measurement date. On the contrary, the first prong is met when “the number of shares that an individual employee is entitled to receive” is determined. APB Opinion 25, ¶ lO.b (emphasis added).

Thus, Treacy’s SEC Form 5 disclosing both quantity and price of the December 9, 1998, option grant is evidence that both prongs of the APB Opinion 25 were set as of February 9, 1999, and it was error to calculate the forfeiture based on a measurement date of April 19, 1999. If the February 9 date were used as the measurement date, the difference in strike price and measurement date price would be smaller, by Treacy’s calculations, resulting in a smaller forfeiture. We therefore vacate this portion of the order of forfeiture and remand for recalculation of the appropriate forfeiture based on the December 9, 1998, option grant.

CONCLUSION

We affirm the judgment of conviction. We vacate the portion of the district court’s order of forfeiture based on the December 9, 1998, option grant and remand for recalculation of the forfeiture amount and entry of a new order of forfeiture order not inconsistent with this opinion.

1 . In addition to relying on the evidence adduced at trial to establish the loss amount, the Government sought to use an analysis by a forensic accounting firm that had been hired by Monster to determine the extent of its losses to the backdating scheme. The district court, however, ruled at sentencing that the outside analysis was not an appropriate basis on which to make factual findings by a preponderance of the evidence, and the court elected to base its findings exclusively on the trial record. On appeal, the Government accepts the approach taken by the district court.

2 . The chart attributed profits to Treacy from six different grants, including the two discussed herein. However, Treacy made clear at oral argument before us that he is appealing the district court’s forfeiture order only with respect to the grants made with the "as of” dates of January 6, 1997, and December 9, 1998. We do not discuss the other four grants, therefore.

3 . Nothing in this holding impugns the force or wisdom of the Justice Department’s internal regulations governing the issuance of subpoenas to journalists, which exist wholly apart from any privilege established in the case law. See 28 C.F.R. § 50.10.