A jury convicted Santiago ("Chago" or "Chagon") Pena-Espinoza of conspiracy to distribute cocaine and distribution of cocaine in violation of 21 U.S.C. Secs. 841(a)(1) and 846. He timely appeals, arguing that the district court erred in admitting into evidence wiretap transcripts of contested accuracy, hearsay testimony summarizing the contents of these transcripts, and statements of one of his alleged co-conspirators. He also contends that the district court erred in its application of relevant United States Sentencing Guidelines provisions. We affirm the conviction and sentence.

During an extensive investigation of a large drug distribution network, FBI agents made over 100 undercover cocaine and heroin purchases. The FBI used telephone monitoring devices to establish the relationship between Pena-Espinoza and the conspiracy. For instance, agents monitored a conversation between Pena-Espinoza and conspirator Arnoldo Aispuro-Aispuro during which Pena-Espinoza agreed to provide Aispuro-Aispuro with cocaine for a sale. Additionally, an undercover agent/purchaser testified that she saw Pena-Espinoza in the room when this sale actually took place, and that the next day Aispuro-Aispuro told her that Pena-Espinoza would supply the drugs for her next order. Agents also surveilled Pena-Espinoza's house and on three separate occasions observed another conspirator entering it immediately before she departed to deliver drugs to undercover agents. Finally, the transcript of a monitored call linked Pena-Espinoza to $22,430 cash found in a seized vehicle belonging to the conspiracy.

A. The Trial.

The trial of Pena-Espinoza and his co-defendant, Mor Rodriguez-Zamora,1 began on June 23, 1992. Recorded telephone conversations constituted a key element of the prosecution's case. These conversations numbered over 4800 in all. Most were in Spanish. During the investigation, the FBI kept a running computerized log noting the participants in each conversation and the substance of key statements. The government prepared actual transcripts for only approximately 50 conversations that it intended to discuss at trial.

Defense counsel knew of the government's trial strategy months in advance. Prosecutors planned to introduce the tapes and transcripts and then put investigating agents on the stand. These agents would serve as "readers" of relevant portions of the transcripts; in some cases they would summarize the contents of conversations as they saw fit. The government turned over to the defendants copies of all tapes as well as a copy of the computer log contents by no later than early December 1991.

Until April 1992, the Federal Public Defender's Office (FPD) was coordinating discovery for approximately twenty defendants in the conspiracy. By April 1992, however, the FPD's defendants had all pleaded out and counsel for Pena-Espinoza took custody of the materials. On the Friday before trial the government provided the actual transcripts--in some cases only rough drafts--that it intended to introduce at trial.

On the first day of trial, just prior to jury selection, counsel for Pena-Espinoza objected to the government's strategy. He admitted later that his failure to object or file a motion in limine until then had been "simply a tactical decision." He also pointed out that the defense had "had no opportunity to verify the accuracy of the transcripts or for that matter the accuracy of the voice identification as to who is purported to be saying what." He did not ask for a continuance of any sort.

The district court overruled Pena-Espinoza's objection. The judge stated that the trial date had been set with the consent of all concerned, that the defendants had had ample warning of the government's planned use of summaries and transcripts and that it was now "too late" to object, and that the trial date had been long delayed precisely to give the defendants time to "get the job done as far as the tapes were concerned."2

The court then inquired as to just how the prosecution wished to present the wiretap evidence. The government responded as follows:

I would simply utilize[,] once the tapes are in[,] readers.... [M]y preference would be not to have the witness read or read the entire conversation, but we can do that if the defendants feel that a summary is not appropriate. We can have the entire conversation read by the reader. I would have the witness identify the exhibit [i.e., the particular transcript] by exhibit number and ask for a summary of that. But if there is an objection, we can have the entire conversation read by the reader.

At oral argument, the government explained that it saw this procedure as a time-saving measure.

However, counsel for Pena-Espinoza immediately responded: "We would object to the summary and require the conversation be read verbatim if that's going to be the procedure." The court then rejected his plan in favor of the government's approach:

I will allow the Government to proceed with the reading by summary as counsel for Mr. Zamora points out, the entire transcript can be used in cross-examination if either counsel feels it is helpful to his cause. In other words, it will be then [ ]at the option of the defense counsel as to whether to require the reading of the full transcript or any portions thereof that are most helpful ...

The trial proceeded in this fashion. The transcripts were not actually admitted until after the testimony of a foundation witness who took the stand after the government's principal witnesses used them. Defense counsel did not object to this sequence.

The jury convicted Pena-Espinoza as charged. He now appeals.

B. Discussion.

Pena-Espinoza makes two arguments with respect to admissibility of the wiretap evidence. He contends first that the district court erred in admitting English-language transcripts of monitored Spanish-language conversations because it did not allow defense counsel a meaningful opportunity to check the transcripts' reliability. Second, he argues that the court erred in permitting the "readers" to summarize in their own words the contents of key transcripts, rather than simply reading the dialogue into evidence.

1. Reliability. Where there is no dispute as to accuracy, we review for abuse of discretion the district court's decision to allow the use of wiretap transcripts during trial and to permit such exhibits into the jury room. United States v. Taghipour, 964 F.2d 908, 910 (9th Cir.), cert. denied, --- U.S. ----, 113 S. Ct. 283, 121 L. Ed. 2d 210 (1992). This standard applies even where the actual tapes are not played for the jury. Id. at 910 n. 1; United States v. Tornabene, 687 F.2d 312, 317 (9th Cir.1982). Moreover, "[a] nonconstitutional evidentiary error will be reversed for an abuse of discretion only if the court's ruling more likely than not affected the verdict." United States v. Yin, 935 F.2d 990, 994 (9th Cir.1991).

In this case, the defense argued that it had had no opportunity to investigate the transcripts' accuracy. On the specific circumstances of this case, however, we hold that abuse of discretion remains the proper standard of review. We find it significant that the defense made no meaningful attempt to check the transcripts for accuracy or at least to obtain a continuance to examine such crucial evidence. See United States v. Booker, 952 F.2d 247, 250 (9th Cir.1991) ("[T]he fact that the transcripts were not available until the day they were to be used does not constitute grounds for reversal. Defense counsel was aware of the existence of the tapes approximately one month before trial and had ample opportunity to review them and to make her own transcripts."). Inasmuch as the vast majority of the taped conversations were in Spanish and they were key evidence, defense counsel had unquestionably known from the outset that English-translation transcripts would be central to the prosecution's case.

These considerations are also highly relevant to our evaluation of Pena-Espinoza's claim on the merits. Moreover, the trial record shows that defense counsel had full opportunities to aggressively cross-examine the "readers" as to transcript contents and for argument to the jury on this subject. See United States v. Armijo, 5 F.3d 1229, 1234 (9th Cir.1993) (focus is on steps taken at trial to ensure accuracy); see also United States v. Carrasco, 887 F.2d 794, 804-07 (7th Cir.1989). In addition, Pena-Espinoza has made no showing, either in the district court or on appeal, that the transcripts were substantially inaccurate. Finally, the court gave the defense's requested jury instruction regarding the transcripts.

We are thus left with largely conclusory allegations of possible inaccuracy and no indication that the court's ruling likely affected the jury's verdict. We therefore hold that the district court did not abuse its discretion in admitting the transcripts and permitting the jury to examine them during deliberations.

2. Summaries by the Readers. Similar concerns inform our analysis of Pena-Espinoza's objection to the district court's ruling allowing the government readers to summarize transcript contents in their own words. We review such an evidentiary ruling for abuse of discretion. United States v. Sanchez-Lopez, 879 F.2d 541, 553 (9th Cir.1989) (hearsay objection).

The court's ruling did not allow the readers to do anything but provide a factual condensation of the transcripts' contents; it did not preclude other forms of objection to the readers' testimony. The transcripts themselves were in evidence and the jury had them to examine during deliberations; the ruling expressly permitted defense counsel to require a reading of the full transcript on cross-examination and to dispute the veracity of the readers' summaries. As a result, such factual condensation, without more, did not likely affect the jury's verdict, and did not constitute reversible error. Yin, 935 F.2d at 994.

As the trial transcripts reveal, the readers actually did much more than provide a factual summary. In particular, Agent Frazier, who was not even qualified as an expert, reviewed transcripts of numerous conversations in which he had not participated and offered his interpretations and opinions as to the significance of many of the participants' statements and thus to the defendants' guilt. Admission of such testimony over timely objection might well have been an abuse of discretion, usurping the jury's function and prejudicing Pena-Espinoza's case.

Unfortunately, on only three occasions did defense counsel object to this aspect of Frazier's testimony (and on two such occasions the court sustained the objection, see RT June 24, 1992 at 188, 191); the general objection to the use of summaries does not extend to preserve this issue with respect to the many other similar instances. See United States v. Gomez-Norena, 908 F.2d 497, 500 (9th Cir.), cert. denied, 498 U.S. 947, 111 S. Ct. 363, 112 L. Ed. 2d 326 (1990). Thus, Pena-Espinoza did not preserve his objection to the readers' editorialization in the course of their summaries. Nor has he raised this specific argument on appeal. Admission of this testimony did not constitute plain error. See United States v. Olano, --- U.S. ----, ---- - ----, 113 S. Ct. 1770, 1776-79, 123 L. Ed. 2d 508 (1993).

Although we do not wish to condone such procedures, in this specific set of circumstances we hold that the district court did not abuse its discretion in permitting the government's readers to summarize transcript contents.

Pena-Espinoza also claims that the district court erred in admitting certain testimony as to the statements of one of his alleged co-conspirators, Arnoldo Aispuro-Aispuro. We review "de novo the legal question of whether the government established a prima facie showing of conspiracy but apply a clearly erroneous standard in reviewing whether a challenged statement was made in the course and furtherance of the conspiracy." United States v. Arias-Villanueva, 998 F.2d 1491, 1502 (9th Cir.) (citation omitted), cert. denied, --- U.S. ----, 114 S. Ct. 573, 126 L. Ed. 2d 472 (1993).3

In particular, Pena-Espinoza argues that the district court improperly admitted the testimony of the investigation's principal undercover officer, Special Agent Sheree Sandlin, regarding her purchase through Aispuro-Aispuro of four ounces of cocaine on March 19, 1991. Sandlin stated that

A: I asked Arnoldo if the price was 850 [dollars per ounce].... He again stated that the price would be 950, and I told him I only had $3400. He then told me that he had to pay the guys, and he indicated the ones in the dining room area by nodding his head in that direction.

Q: Was Mr. Pena-Espinoza still sitting at the dining room table?

A: Yes, he was.

....

Q: Did you have some further discussion at that point about where the stuff from the 19th and the 20th would come from?

A: Yes. I asked Arnoldo where the stuff came from, referring to the cocaine that I bought yesterday, and he said Chago. He also told me that the stuff or cocaine that I would be receiving that day would be coming from Chago.

"Chago" had been elsewhere identified as a nickname for Santiago Pena-Espinoza.

On the basis of the other evidence in the record, principally the wiretap transcripts, a drug courier's several visits to Pena-Espinoza's house just prior to making deliveries, and the defendant's presence during one of Sandlin's drug purchases, we hold that the district court did not err in finding that a drug distribution conspiracy existed and did not clearly err in finding that Aispuro-Aispuro's statements to Sandlin were made in the course and furtherance of the conspiracy. As a result, the district court did not abuse its discretion in admitting Sandlin's testimony over the defense's objection.

Finally, Pena-Espinoza challenges the district court's application of the United States Sentencing Guidelines to his conspiracy conviction. In calculating the amount of cocaine involved in Pena-Espinoza's activity pursuant to Section 1B1.3(a)(1)(B) of the Guidelines (Nov. 1991 ed.), which defines conduct relevant to sentencing to include all harm resulting from the defendant's criminal activity, the court imputed an amount based upon the $22,430 in cash seized from a vehicle used by the conspiracy. We review for clear error the district court's factual findings in the sentencing phase. United States v. Chapnick, 963 F.2d 224, 226 (9th Cir.1992).

Pena-Espinoza contends that the government failed to demonstrate that he either knew of or could reasonably have foreseen this amount of money or the transactions that generated it. The district court relied primarily on the transcript of a phone conversation between Arnoldo Aispuro-Aispuro and Pena-Espinoza. In this March 2, 1991 conversation, the two discussed the seizure of the vehicle in question, which had occurred the day before. The driver of the seized car, co-defendant Marcos Aispuro-Pena, was Pena-Espinoza's nephew, and he was at Pena-Espinoza's residence on March 2nd. Surveillance agents had seen Marcos Aispuro and Pena-Espinoza together in this car only days before.

The conversation, in pertinent part, went as follows:

Aispuro-Aispuro: Did they take the car away?

Pena-Espinoza: Everything down to the underwear.

Aispuro-Aispuro: Really?

Pena-Espinoza: Yes!Aispuro-Aispuro: You went, to pick him up?

Pena-Espinoza: No, he came in a taxi.

...

Aispuro-Aispuro: Will you get the car out?

Pena-Espinoza: Who knows?

Aispuro-Aispuro: How much did they take away?

Pena-Espinoza: That.

Aispuro-Aispuro: Twenty-five?

Pena-Espinoza: Three?

Aispuro-Aispuro: Huh?

Pena-Espinoza: Twenty-three.

Aispuro-Aispuro: Twenty-three?

Pena-Espinoza: Yes.

Aispuro-Aispuro: Damn! Are you going to get it back?

Pena-Espinoza: What?

Aispuro-Aispuro: Are you going to fight it?

Pena-Espinoza: Oh, I don't know, I don't know, what do you think?

Pena-Espinoza argues that this conversation was ambiguous, that the Spanish pronoun translated as "you" could equally well have meant "they". However, in the context of the other evidence and the specifics of the dialogue from which this excerpt is taken, we hold that the district court did not clearly err in finding that the cash in the car was related to drug transactions by the conspiracy of which Pena-Espinoza was a part, that he knew of such transactions, and that he could readily have foreseen them.

We thus AFFIRM the appellant's convictions and sentence.

The district court allowed testimony by law enforcement agents in which they summarized critical wire-tapped conversations to which they were not a party. Such a practice is both unprecedented and indefensible. Accordingly, I dissent.

I note at the outset that the majority rests its affirmance on a narrow view of what this case requires us to decide. The majority concludes that the district court's ruling only authorized the government to introduce "factual condensation[s] of the transcripts' contents." Ante at 360. Therefore, according to the majority, the defendant's objection to the ruling preserved only the issue whether "factual condensations" were admissible. The majority then concludes that the defendant never properly objected to the non-factual testimony offered by the agents.

As a result of its narrow construction of the district judge's ruling and the defendant's objection, the majority reviews the admission of different portions of the agents' testimony under different standards of review. It reviews the admission of the agents' "factual condensations" for an abuse of discretion. In contrast, it reviews the remainder of the agents' non-factual testimony only for plain error.

I believe that the majority's interpretation of the district court's ruling places the vast bulk of the agents' testimony beyond the scope of the ruling. The majority itself suggests as much when it states that the agents' summaries "actually did much more than provide a factual summary." Ante at 360. Unsurprisingly, a review of the agents' testimony confirms that only a small portion of it can fairly be characterized as consisting of factual summations of the transcripts.

During most of their testimony, the agents purported to summarize the conversations directly. At those points, the agents were not simply providing factual summations of the transcripts. Instead, they were describing conversations about which they had no factual knowledge based solely on their opinions that the transcripts were accurate. It does not appear that the majority means to deem such unqualified opinion testimony to be within the scope of its construction of the district court's ruling.

More important, throughout their testimony the agents made subjective assessments of the relevance of certain conversations. In particular, when Agent Frazier informed the jury of the "pertinent" parts of the conversations, he acted as far more than a reporter or condenser of the contents of the transcripts. Instead, he played the role of commentator. Analysis of the pertinence of evidence is a function of jurors, not witnesses. Although the majority does not define what it means by "factual condensations", relevancy determinations clearly fall within the majority's description of the non-factual testimony that it concedes "might well have been an abuse of discretion, usurping the jury's function and prejudicing Pena-Espinoza's case." Ante at 360.

Because under its construction of the district court's ruling most of the agents' testimony must be deemed to have been admitted without objection, the majority decides only that the admission of most of the summary testimony did not constitute plain error. It also holds that the admission of the small amount of testimony that could be called "factual condensations" had no effect on the verdict in this case and thus did not constitute reversible error.

These narrow holdings permit the majority to uphold Pena-Espinoza's conviction. However, the court's opinion does not answer the more general question whether a district court abuses its discretion when it permits law enforcement agents to summarize conversations to which they were not a party.

Despite the limited nature of the majority's holding, I am compelled to dissent. My disagreement stems from my conclusion that the majority's highly limited view of the scope of the district court's ruling is improper. To me, it is clear that the ruling in question was not meant to be limited to the mechanical rendering of excerpts from the transcripts; certainly, the defendant's objection was not. He intended to object to the entire process.

The district court itself never used the phrase "factual condensations" in describing the scope of its ruling. That restrictive description of the ruling is one wholly of the majority's creation. It is premised on the assumption that there exists some species of summaries that consist merely of "factual condensations". However, the majority's categorization of the evidence is both artificial and highly unrealistic. Summaries are never simply factual condensations. They necessarily reflect "interpretations and opinions" of the evidence being summarized. Ante at 360. When the district court authorized the agents to summarize the "contents" of the transcripts, it unquestionably permitted them to introduce the subjective descriptions and characterizations of the conversations to which they testified.

Thus, I believe that the majority is wrong to review the bulk of the agents' testimony under the plain error standard of review. Although in my view the admission of the agents' testimony would, if not objected to, constitute plain error, I believe that the defendant did object to the introduction of all the testimony when he objected to the district court's ruling, and that in fact the ruling applied to the testimony in its entirety. Accordingly, I explain below why I conclude that the district court abused its discretion in permitting the agents to testify as they did.

The plain fact is that there is no precedent for the admission of summaries such as those used in this case. The lack of precedent is hardly surprising. We have long recognized that tape recordings constitute the "actual evidence of the conversations." United States v. Tornabene, 687 F.2d 312, 317 (9th Cir.1982); see also Federal Rules of Evidence 1002 (stating that "[t]o prove the content of a writing, recording, or photograph, the original writing, recording, or photograph is required"). As a result, we have held that even word-for-word transcripts are admissible only under certain conditions. United States v. Taghipour, 964 F.2d 908, 910 (9th Cir.), cert. denied, --- U.S. ----, 113 S. Ct. 283, 121 L. Ed. 2d 210 (1992); United States v. Booker, 952 F.2d 247, 250 (9th Cir.1991); Tornabene, 687 F.2d at 317.

Before admitting transcripts into evidence, we have considered both whether the trial judge has reviewed them for accuracy, Booker, 952 F.2d at 250, and whether the parties have stipulated to their fidelity to the tapes, Taghipour, 964 F.2d at 910. We have also looked at whether a government agent who participated in the taped conversation has testified to the reliability of the transcripts. United States v. Gonzales-Benitez, 537 F.2d 1051, 1054 (9th Cir.1976) (holding that in the absence of the actual tapes, "testimony by the participants ... [may be] sufficient to establish what was said.") (emphasis added).

Moreover, we have noted the importance of reminding the jurors of the primacy of the tapes, United States v. Armijo, 5 F.3d 1229, 1234-35 (9th Cir.1993); United States v. Hsieh Hui Mei Chen, 754 F.2d 817, 824 (9th Cir.1985) (same), and we have suggested that a judge should instruct the jury to resolve discrepancies between the tape and the transcript in favor of the tape, see Booker, 952 F.2d at 249. We have noted the value of such limiting instructions even when the taped conversations took place in Farsi, a language incomprehensible to most jurors. See Taghipour, 964 F.2d at 910 & n. 2.

Given the safeguards that we have delineated for the admission of transcripts of tapes, I do not understand how summaries of transcripts could be admitted without similar constraints. Verbatim transcripts are likely to provide far more accurate renderings of conversations than are summaries. Moreover, as the childhood game of "telephone" well demonstrates, words change significantly in the course of their re-telling by third parties. There is simply no way to summarize the contents of a transcript without offering to some degree a subjective view of their meaning and import. Indeed, we have previously noted that when a witness summarizes the underlying evidence it "lends that witness' credibility to that evidence and may obscure the jury's original evaluation of" it. United States v. Baker, 10 F.3d 1374, 1412 (9th Cir.1993). For that reason, I do not believe that district courts have the discretion to admit testimony of the type at issue here. Nor, for the reasons I have explained, do I read the majority opinion as holding that they do.

The form that the summaries took in this case demonstrates the problems that result from admitting testimony in the form of summaries of recorded telephone conversations. Each witness testified as if his testimony were based on personal knowledge, even though it was not. Thus, each agent assumed the dual role of fact and opinion expert. "[T]his intermingling of the two roles increased the possibility that the jurors would not distinguish" between those statements that were opinion and those that were not. United States v. De Soto, 885 F.2d 354, 360 (7th Cir.1989) (describing the risks presented when a police officer testifies as both an expert and an eyewitness).

The fact that neither witness had ever been qualified to assume either role exacerbated the problems engendered by the dual nature of their summary testimony. The witnesses described conclusively what words were used in the conversations and which parties were speaking when. The government introduced the witnesses' testimony without first showing that either one had listened to the tapes, recognized the voices on them, or played any role in preparing the transcripts. United States v. Rengifo, 789 F.2d 975, 983 (1st Cir.1986) (describing requirements for witness to qualify as an expert in translation of recordings). Moreover, Agent Frazier characterized certain words as being drug-related despite the lack of any foundation for such testimony. Cf. United States v. Borrone-Iglar, 468 F.2d 419, 420-21 (2nd Cir.1972) (considering whether witness was qualified to characterize the meaning of narcotics vernacular). In short, the law enforcement witnesses here did not "perform an essentially mechanical task" in providing the summaries. Rengifo, 789 F.2d at 983. Instead, I believe that their testimony "change[d] the essential nature of the transcript evidence" by enhancing its credibility and altering its meaning. Id.

I note finally that I disagree with the majority as to whether the admission of the objected-to testimony constituted reversible error. I think that we must examine each agent's testimony in its entirety. There is simply no way to isolate "factual condensations". Even if there were, none was suggested here. The admission of the agents' testimony clearly prejudiced the defendant. The "summaries" provided the jurors with unqualified but seemingly authoritative opinion as to the contents, meaning, and significance of crucial conversations. Given the large number of transcripts involved, it is clear that the testimony provided the jurors with the most readily accessible description of the conversations. Because I believe it is likely that the jurors relied on the summaries in concluding that the conversations were incriminating, and because I believe that the district court abused its discretion in admitting those summaries, I conclude that we are required to reverse the convictions. Accordingly, I dissent.