Dispatches From The Martha Stewart Trial

Reasonable Doubt

Chance of conviction
Wednesday, Feb. 11: 32 percent
The defense created doubt about the accuracy of the alleged false statements made in the defendants’ unsworn, unrecorded, and untranscribed interviews. Stewart’s chances of conviction drop to 32 percent.

In the conclusion of his cross-examination of SEC attorney Helene Glotzer this morning, Richard Strassberg all but vaporized one element of the “false statement” charge against Peter Bacanovic—a statement Bacanovic allegedly made in his unsworn, unrecorded, and untranscribed interview with the SEC on Jan. 7, 2002. Later, cross-examinations by Stewart attorneys Bob Morvillo and Jack Tigue went a long way toward crippling the two similar false-statement charges against Martha Stewart.

To review: Three of the nine counts in United States v. Martha Stewart and Peter Bacanovic are “false statement” charges based on statements that Stewart and Bacanovic allegedly made in voluntary interviews with federal investigators. Three of the other counts, the two “obstruction of justice” charges and the single “conspiracy” charge, also depend in part on these statements. As I described in my pre-trial analysis, the interviews in which the statements were (allegedly) made were neither recorded or transcribed: The only record of them is hand-written notes. As a result, I expected that, before challenging the alleged falsity of the statements, the defense might question whether the defendants even made them. This effort began in earnest today.

In rapid, aggressive questioning, Strassberg hammered away at Helene Glotzer, an SEC attorney who participated in the interviews. Strassberg reminded Glotzer (and the jury) of her prior testimony that, on Jan. 3, 2002, Douglas Faneuil said he took Stewart’s sell order. Then Strassberg mentioned Glotzer’s testimony that, four days later, on Jan. 7, Bacanovic said that he took Stewart’s order. Lastly, Strassberg reminded everyone that, upon hearing this, Glotzer made no attempt to clarify the discrepancy. The unasked question reverberated: Did this make sense? No. And this attack was only the beginning.

Yesterday, Strassberg got Glotzer to say that she used the “royal we” to refer to SEC activities—”We are doing this; we are doing that”—and she confirmed that this didn’t mean that she was doing those things. Similarly, he suggested, in Peter Bacanovic’s Jan. 7 interview with the SEC, Bacanovic had used “we” to describe the activities of his Merrill Lynch team and not himself—and that, with regard to the alleged false statement, the SEC had misunderstood him. Specifically, Strassberg suggested, Bacanovic had said that “we” had received a call from Martha Stewart on Dec. 27, 2001, and the SEC had mistakenly interpreted this to mean that he had received the call. (This alleged false statement is key to not only Count 2 of the indictment, “False Statements By Peter Bacanovic,” but Count 1, the conspiracy charge: The government is arguing that Bacanovic and Stewart allegedly conspired to make this statement to keep the investigators’ focus off Faneuil.)

Yesterday, in support of his argument, Strassberg asked to play a tape of Bacanovic’s Feb. 13 testimony at the SEC—the only interview in the entire investigation that was sworn, recorded, and transcribed. In this interview, Strassberg said, the SEC made the same “We/I” mistake even though the interview was recorded. After several feisty arguments with Judge Cedarbaum (in which Cedarbaum, missing the point, dressed Strassberg down for not requesting that the transcript be corrected months ago), Strassberg finally got to play his tape.

At Strassberg’s direction, Helene Glotzer donned headphones and listened to a tiny portion of Bacanovic’s SEC testimony. As she listened, she compared what she heard to the SEC’s formal transcript of the testimony. As Strassberg had already reminded everyone, the transcript Glotzer was reading had been vetted by not only the court reporter who created it, but Glotzer herself, who had listened to the tape, compared it to the transcript, and then approved the transcript for submission in this trial. When Glotzer finished listening, Strassberg asked her to explain the mistakes she had just heard to the jury.

According to the transcript, Glotzer testified, the investigators asked Bacanovic the following question [all italicized testimony is paraphrased]:

You received a call from who on Dec. 27th, 2001?

To which he replied:

I got a call from Alan Goldberg [Sam Waksal’s accountant]…

On the transcript, in other words, it was clear that Bacanovic had spoken to Goldberg. On the tape, however—the tape against which the transcript had supposedly been checked by Glotzer and others—it was not at all clear:

Q: You received a call from who on Dec. 27th, 2001?A: Well, the calls began from Alan Goldberg …

Now, as far as records of oral conversations go, court transcriptions are thought to be pretty accurate—and, on a percentage basis, they are. Court reporters still make mistakes, however, and sometimes the mistakes are crucial. In the case of Bacanovic’s Feb. 13 SEC testimony, moreover, the court reporter’s transcript was subsequently scrutinized by a trained, experienced SEC attorney, one presumably involved in the decision to charge Bacanovic with perjury (a felony punishable by more than a year in jail). As Strassberg’s demonstration revealed, however, even this court transcript contained potentially crucial mistakes.

So, now we go back to the Jan. 7 Bacanovic interview. The only record of this interview, as with the two interviews in which Martha Stewart allegedly made false statements, are handwritten notes. In the case of the Bacanovic interview, one of the two alleged false statements is that “he had a conversation with Martha Stewart in which he told Stewart that ImClone’s price had dropped and Stewart told him to sell her ImClone stock.” If Strassberg’s little tape demonstration didn’t convince the jury that there is reasonable doubt about whether Bacanovic made this statement, nothing will.

When Strassberg finished, Bob Morvillo had his turn with Glotzer. For obvious reasons, the unsworn, unrecorded, and untranscribed interviews he focused on were Stewart’s.

Q: Ms. Glotzer, you testified about the substance of Ms. Stewart’s interview at the U.S. attorney’s office on Feb. 4. Prior to your testimony, did you do anything to refresh your recollection?A: Yes.Q: How did you refresh your recollection?A: I read the 302 by FBI Special Agent Farmer.

[A “302” is a typed summary of the hand-written notes taken during the interview.]

Q: How long was it?A: Ten pages.Q: Did you read it in its entirety?A: Yes. Q: When was the last time you read it?A: This week.Q: How many times have you read the 302?A: Since the interview? I’ve probably read it 10 times at least.Q: In preparing for your testimony here, how many times have you read it?A: Five to six times.

Morvillo’s point, of course, was that when you are trying to recall a group conversation of two years ago, and you have a typed summary of that conversation that you have read at least 10 times since the conversation took place (including three times in the last month, according to additional Glotzer testimony), what is it that you are really “recalling”—the conversation, or the typed summary?

And that brought us to FBI Agent Catherine Farmer, the only government representative who took notes in the two Stewart interviews and the one who created the 302 that Glotzer, Farmer, and others have used to “refresh their recollections” about the interviews. On direct examination, Special Agent Farmer described—with striking similarities—what Helene Glotzer had described the day before: what Stewart had said in her two interviews. Then Stewart attorney Jack Tigue cross-examined her.

Q: Of all the ways of recording conversations, Ms. Farmer, would you say that hand-written notes is the least reliable?A: Certainly.Q: It is difficult to capture everything that is said?A: Yes.

Tigue asked Farmer how long Stewart’s first interview had lasted. Approximately two hours, Farmer said. Tigue asked how long Peter Bacanovic’s Feb. 13 testimony at the SEC had lasted. Two hours and 50 minutes. This interview, Tigue continued, had produced a transcript that was 163 pages long. How long were Farmer’s notes of first Stewart interview? Twenty-five pages. How long was the 302 of Stewart’s interview? Nine pages. In Bacanovic’s SEC testimony, Tigue added, more than 900 questions had been asked (and recorded). In the Stewart interviews, meanwhile, Farmer hadn’t taken any notes of the questions asked—only Stewart’s answers.

Q: Would you agree that knowing questions is critical to understanding answers?A: Yes. …Q: These notes make no attempt to be verbatim? A: That’s correct.Q: You are not able to take notes that read verbatim?A: No.

To be fair, the “false statement” charges against Stewart involve many more specifications and alleged statements than the single charge against Bacanovic, and it is highly unlikely that the government has gotten them all wrong. This said, based on today’s events, I think it is safe to say that no investigators, no matter how professional, careful, and/or well-intentioned, will ever be sure that oral statements recorded only via hand-written notes accurately reflect the statements themselves. And if you’re going to charge someone with a felony and send him or her to prison for making such statements, it seems to me, you ought to be sure.