PER CURIAM.
Defendants, Franldyn Rivera-Santiago and Edwin Alamo-Silva, challenge their convictions and sentences following a jury trial. For the. reasons discussed below, we vacate their convictions and remand for a new trial.
I.
Background
We recount only those facts necessary to resolve the instant appeals. On January 25, 1995, a grand jury returned a three-count indictment against the defendants, charging them with aiding and abetting each other in attempting to possess with intent to distribute • narcotics in violation of 21 U.S.C. §§ 841(a)(1), 943 and 18 U.S.C. § 2; aiding and abetting- each other in attempting to import narcotics into the United' States in violation of 21 U.S.C. §§ 952(a), 963 and 18 U.S.C. § 2; and aiding and abetting each other' in attempting to possess on board a vessel of the United States with intent to distribute narcotics in violation of 46 U.S.C.App. § 1903(a), (b)(2)(c), (f), (j) and 18 U.S.C. § 2.
At trial, the government elicited testimony from U.S. Customs Service air interdiction officers who were assigned to three aircraft operating off the southeast coast of Puerto Rico on the night of January 4,1995, and the morning of January 5, 1995. The officers were investigating what was perceived to be a suspicious aircraft flying with its lights off and without a flight plan from South America toward Puerto Rico. The government’s first witness was Leslie Robb, who operated the radar and Forward Looking Infrared camera (“FLIR”)
on one of the aircraft, Omaha 42. Robb testified that after circling for nearly forty minutes, the suspicious aircraft dropped several objects into the water at 12:39 a.m. at a point approximately six miles off the coast of Patillas, Puerto Rico. The splashes were captured on a videotape of Omaha 42’s FLIR, which was submitted to the jury as evidence along with videotapes of the FLIRs from the other two aircraft. The videotapes included the radio communications among the air interdiction officers and other law enforcement personnel that occurred contemporaneously with the images produced by the FLIRs. These radio communications revealed that immediately before the airdrop occurred, a vessel was seen flashing its lights in the area near the suspicious aircraft, and that the aircraft which had turned its lights on at some point before the airdrop turned them off shortly thereafter.
The government’s second witness, Raul Antonio Rivera-Calleja (“Rivera”), operated the radar and FLIR systems on a second plane, Omaha 02, which began looking for marine targets after the airdrop had occurred. At approximately 1:04 a.m., Rivera acquired a vessel on radar approximately two miles from the site of the airdrop. This target was the only one that Rivera was able to locate on his radar, which covered a twenty-five-mile radius around the aircraft. Rivera testified, that after he had located the vessel on radar, his fellow crew members informed him that they saw no lights on the surface of the water, and that the vessel’s navigation lights had to have been turned off. Rivera further testified that shortly after acquiring the vessel on radar, his FLIR showed two objects floating in the water approximately twenty feet from the rear of the vessel.
The government’s next witness, David Cruciger, was the pilot in command of Omaha 02. Cruciger offered the following de
scription of the airdrop on direct examination:
Once again while flying in this pattern, we received information from Omaha .42 that a drop was taMng place and that they were seeing the splashes. Chris Thorton [the co-pilot aboard Omaha 02] said it, directed my attention out the right-hand window of the aircraft. I banked the aircraft over so I could see out in that direction; and with the aircraft banked, I could see down into the water a flashing light. It was described by Omaha 42. As we watched the light, I saw an aircraft or what I believed to be an aircraft turn on its navigation recognition lights and fly at low altitude over the lights that were flashing in the water.
When asked during cross-examination whether he had seen the lights before the drop occurred, Cruciger stated that he “believe[d] it was during the drop.”
The government’s fourth witness, John Al-pers, operated the radar and FLIR system on the third aircraft, Omaha 38. Alpers testified that he located a vessel — the same one next to which Rivera’s FLIR would later detect two objects floating in the water — on radar at approximately 12:45 a.m., approximately two miles from the site of the airdrop, and that, according to the co-pilot aboard Omaha 38, all of the vessel’s lights were out. Alpers also testified that although he located one object on his radar that he believed to be a reef, he located no other vessels on his twenty-five-mile radar.
The government’s witnesses testified that the vessel remained stationary from the time Alpers acquired it on his FLIR at 12:54 a.m., approximately three miles from the site of the airdrop, until a coast guard helicopter arrived at approximately 1:13 a.m. and shined a bright light on it. They further testified that the vessel began to move toward shore after the coast guard helicopter illuminated it, and that the defendants were found aboard the vessel when it arrived on shore and were promptly arrested. Although no contraband was found aboard the defendants’ vessel, the government introduced into evidence four bales of cocaine that were found floating in the water in the vicinity of the drop site. The first bale was found at approximately 2:00 a.m. on January 5, 1996, and three more were found at approximately 1:30 p.m. the same day tied together with rope of a type found on the defendants’ vessel. Two of the government’s witnesses speculated that, following the airdrop, the defendants had gathered the three bales that were found tied together and, rather than bringing the bales aboard, placed them in tow so as to facilitate disposal in . the event the defendants’ scheme was discovered.
The government also introduced into evidence a business card found in defendant Rivera-Santiago’s wallet, which was discovered in his car, bearing the coordinates of a spot approximately seven miles away from the site of the airdrop. In addition, the government’s witnesses testified that they found defendant Alamo-Silva’s Toyota 4-Runner at his girlfriend’s house, which was accessible via a pathway to the beach and was located approximately one-quarter of a mile away from the point where the defendants arrived on shore. The keys to the 4-Runner were found in the vehicle, which was parked with its back to the water and its rear seats folded down. Finally, the government’s evidence indicated that the registration number painted on the back of the vessel on which the defendants were found, which belonged to defendant Rivera-Santiago, differed by two letters from the number under which the vessel had been registered with the Puerto Rico Department of Natural Resources.
The defendants contended that they were out fishing on the night of the airdrop and were simply caught in the wrong place at the wrong time. They elicited testimony concerning other vessels that might have been in the area at the time of the airdrop but were not detected on radar, including a vessel found abandoned on shore on the morning of January 5, 1995, that contained marijuana residue. Rivera-Santiago testified that his
vessel was operating without navigation lights on the night in question because the ■vessel’s built-in navigation lights had been damaged on a previous occasion and because he had stopped using the portable navigation light he had recently purchased for the boat. He explained that the portable navigation light “wouldn’t stay fixed for a long time,” would fall every time the boat jumped, and “didn’t work.”
His testimony also indicated that he' began fixing one of the engines on the boat at approximately 11:00 p.m. with only the aid of a flashlight, and that the vessel was already moving when the coast guard helicopter arrived at 1:13 a.m. Finally, the defendants elicited testimony concerning the maximum speed of Rivera-Santiago’s vessel, suggesting that it might have been difficult for the boat to retrieve at least three bales of cocaine and travel approximately two miles, as the government’s evidence suggested, during the six-minute interval between the airdrop and the time Alpers detected Rivera-Santiago’s vessel on radar. They also argued this point to the jury.
The juty began deliberating late in the afternoon on April 10, 1995, and, at 7:50 p.m. the following day, informed the trial judge that it was unable to reach a verdict but wanted to come back the next day to continue its deliberations. The next morning, the jury submitted the following message to the court:
We wish to obtain the following information from the transcription notes to clarify some doubts:
1. The first time the suspect air craft was detected in the fishing area (the hour)
2. The time of the air drop
3. The time the suspect vessel was detected in the fishing area
4.If there were any sign of flashing lights from the suspect aircraft and suspect vessel.
After trial counsel for defendant Alamo-Silva unsuccessfully argued that providing answers to any of the jury’s questions would invade the province of the jury, the parties and the court agreed on the responses to the first three questions. In response to the fourth question, and at the government’s request, the trial judge elected to read back to the jury part of David Cruciger’s testimony. This decision came over the objection of counsel for each of the defendants. Noting that Cruciger was the only witness who testified that he saw flashing lights,
defense counsel asked that testimony from the witnesses who did not see flashing lights be read into evidence, and further argued that it was not clear what the jury meant by “suspect vessel.”
The trial judge addressed the jury as follows:
I have your, four questions. Let me say that the answer to two of your questions you have to see the [videotape of the] FLIR; you should see the FLIR. But I’m going to answer two questions. The first — your first question is the first time the suspect aircraft was detected in the fishing area, the hour. That you may look into the FLIRs.
Then the time of the airdrops, there’s stipulation that it’s 12:39 a.m., 12:39 a.m., time of the airdrop.
The third question, the time the suspect vessel was detected in the fishing area, that you have to look it up in the — from the FLIR tapes.
Fourth — I’m going to answer that question now — if there — if there were any sign
of flashing lights from the suspect aircraft and the suspect vessel, I’m going to read you the testimony of David Crueiger. Listen carefully. [The court reporter then read back the following portion of Cruciger’s testimony.]:
Once again while flying in this pattern, we received information from Omaha 42 that a drop was taking place and that they were seeing the splashes. Chris Thorton, Officer Thorton said it, directed my attention out the right-hand window of the aircraft. I banked the aircraft over so I could see out in that direction; and with the aircraft banked, I could see down into the water a flashing light. It was described by Omaha 42. As we watched the light, I saw an aircraft or what I believed to be an aircraft turn on its navigation recognition lights and fly at low altitude over the lights that were flashing in the water.
[The trial judge continued.] Okay. Very well. You may go back to your deliberations.
The jury returned guilty verdicts against the defendants on all counts approximately two hours after the trial judge answered its questions, apparently having eaten lunch in the interim. Defendant Alamo-Silva was sentenced to prison for 295 months. Defendant Rivera-Santiago was sentenced to prison for life.
II.
Discussion
A.
The Trial Judge’s Response to the Jury’s Fourth Question
We first address the defendants’ contention that the trial judge committed reversible error when, in response to the jury’s fourth question, he selected a portion of Cruciger’s testimony and had it read to the jury.
The Sixth Amendment guarantees a defendant in a criminal ease the right to a trial by jury. We have previously noted that “[u]ndeniably inherent in the constitutional guarantee of trial by jury is the principle that a court may not step in and direct a finding of contested fact in favor of the prosecution ‘regardless of how overwhelmingly the evidence may point in that direction.’”
United States v. Argentine,
814 F.2d 783, 788 (1st Cir.1987) (quoting
United States v. Martin Linen Supply Co.,
430 U.S. 564, 573, 97 S.Ct. 1349, 1355-56, 51 L.Ed.2d 642 (1977)). Although the district court may, at its discretion, reread testimony where the jury makes a request to have specific testimony reread,
see, e.g., United States v. Bennett,
75 F.3d 40, 46 (1st Cir.),
cert. denied,
— U.S.—, 117 S.Ct. 130, 136 L.Ed.2d 79 (1996);
United States v. Aubin,
961 F.2d 980, 983-84 (1st Cir.),
cert. denied,
506 U.S. 886, 113 S.Ct. 248, 121 L.Ed.2d 180 (1992), we have noted that the culling of testimony in response to a jury’s open-ended question may, “in effect, make the court a finder of fact,”
see Aubin,
961 F.2d at 983 (quoting
United States v. Almonte,
594 F.2d 261, 265 (1st Cir.1979)), and have found constitutional error where a district court’s answer to a jury’s factual question had the effect of mandating that the jury reach a conclusion on a particular issue.
Argentine,
814 F.2d at 787-88.
Our analysis of the trial judge’s answer to the jury’s fourth question in the context of the evidence elicited during the course of the trial compels us to conclude that the trial judge usurped the jury’s fact-finding role as to the subject matter of that question, and, in so doing, deprived the defendants of their right to trial by jury. In reaching this conclusion, we note that two of the “vices” we identified in
Argentine
are not present here. First, the trial judge did not expressly represent that the parties had reached an agreement as to the subject matter of the jury’s question. Second, rather than presenting his answer to the jury’s question as “accomplished fact,” the trial judge informed the jury that the evidence he was recounting was the testimony of a particular witness.
See id.
at 787. However, it is evident from a review of the record that the substance of the court’s answer together with the context in which it was delivered brought about the same prohibited result that we found in
Argentine
for three reasons.
First, the trial judge selected only a part of Cruciger’s testimony given on direct examination to be read in response to the jury’s question and, in so doing, necessarily sug
gested to the jury that this testimony would provide “the” answer to the jury’s question. This suggestion had the effect of both encouraging the jury to believe Crueiger and discouraging the jury from considering and possibly crediting alternative accounts of the events surrounding the airdrop. The record contains evidence that was inconsistent with, if not contradictory to, Cruciger’s assertion that he saw an exchange of lights during the airdrop between the suspicious aircraft and an object in the water. For example, Cru-ciger’s testimony was at odds with Rivera-Santiago’s assertions that his vessel’s built-in navigation lights were not working and that he had stopped using the portable navigation light that he had recently purchased. In addition, Crueiger testified on direct examination that he began looking for flashes of light from sea level
after
the suspect aircraft had begun dropping objects into the water and then stated on cross-examination that he believed he saw lights during the airdrop. However, the audio portion of the videotape indicates that unidentified air interdiction officers viewed flashing lights
before
any objects were dropped, and contains no mention of lights from sea level during or after the airdrop. At a minimdm, these inconsistencies raise questions about when or whether Crueiger saw flashing lights from sea level and, if he saw lights, whether those lights came from Rivera-Santiago’s vessel or from another vessel.
Second, the context in which the trial judge gave his response to the jury’s fourth question had the effect of placing his imprimatur on the facts contained in that portion of Cruciger’s testimony that was read to the jury. It is significant that the jury was posing questions in order to obtain information from the record “to clarify some doubts.” In answering the first three questions, the trial judge either provided an unequivocal statement of fact to which the parties had stipulated (question no. 2) or directed the jury to the FLIR videotapes where the answer to their questions could be found (question nos. 1 & 3). However, in responding to the jury’s fourth question the trial judge began by stating “I’m going to answer that question now,” and then had part of Cruciger’s testimony read. The net effect of what the trial judge did was to focus the jury’s attention on only part of Cruciger’s testimony concerning flashing lights and away from other evidence given by Crueiger and others that was relevant to a resolution of the doubts the jury expressed in its note about the existence of flashing lights from the air and from sea level.
Finally, by referring the jury to Crueiger’s testimony, the trial judge suggested to the jury that the “suspect vessel” (the term used in the jury’s question and throughout the trial to describe the boat on which the defendants were found), and not another vessel, was the source of the light that Crueiger claimed to have seen. Although, as noted above, Rivera and Alpers testified that Rivera-Santiago’s vessel was the only one that showed up on radar at the time of the airdrop, the defendants presented evidence suggesting that there were other vessels in the area. The trial judge’s answer to the question confirmed an assumption inherent in the jury’s question, i.e., that the vessel seen flashing its lights was the same “suspect vessel” that Alpers picked up on radar six minutes after the airdrop occurred and approximately two miles away.
For the foregoing reasons, we conclude that the trial judge’s response to the jury’s fourth question invaded the province of the jury.
We must next turn our attention to the question of whether the district court’s error can be termed “harmless.” As we noted in
Argentine,
in cases involving errors
of constitutional dimension the harmless-error inquiry focuses on the existence of a reasonable possibility that the error at issue influenced the jury in reaching the verdict.
Argentine,
814 F.2d at 789 (citing
Fahy v. Connecticut,
375 U.S. 85, 86-87, 84 S.Ct. 229, 230-31, 11 L.Ed.2d 171 (1963)). Phrased another way, we must now determine whether the constitutional error was harmless beyond a reasonable doubt.
Id.
at 789;
see also Sullivan v. Louisiana,
508 U.S. 275, 279, 113 S.Ct. 2078, 2081-82, 124 L.Ed.2d 182 (1993);
Chapman v. California,
386 U.S. 18, 23-24, 87 S.Ct. 824, 827-28, 17 L.Ed.2d 705 (1967);
United States v. Trenkler,
61 F.3d 45, 60 n. 22 (1st Cir.1995).
As we have previously noted, the jury in its message to the trial judge sought specific factual information “to clarify some doubts.” Therefore, we can reasonably infer from the context in which the jury asked the fourth question that some or all of the jurors, after deliberating for a little more than one day, entertained “some doubts” about the existence of flashing lights from the “suspect aircraft” and “suspect vessel.” That the jury had doubts about this issue is significant for two reasons. First, the existence or nonexistence of flashing lights, their location, their source, and their timing are matters that were central both to the government’s and the defendants’ theories of the ease.
Second, the manner in which the trial judge culled the evidence effectively determined the outcome of how the jurors would resolve their doubts since they were directed to only part of the evidence concerning lights (i.e., a portion of the testimony of the only witness who testified that he saw lights) instead of being instructed to consider and weigh all of the evidence relating to that issue adduced at trial.
The defendants were entitled to have their theory of the case, as developed through their evidence, presented to the jury on an equal footing with the government’s theory of the case. This did not occur because the trial judge’s response tipped the scales in favor of the government’s theory.
We also note that the jury had expressed to the trial judge an inability to reach a verdict the night before posing its questions and arrived at guilty verdicts two hours after receiving the answers, raising an inference that the trial judge’s response influenced the verdicts. While such an inference in and of itself is not controlling in our evaluation of harmless error, it is a factor that can be weighed along with other factors.
We have considered the trial judge’s error and its effects, as previously discussed, in the context of the entire trial record. The government presented significant circumstantial evidence pointing to the defendants’ guilt. However, under the applicable standard the verdicts can stand only if we find that the error was harmless beyond a reasonable doubt. We cannot make such a finding. The government’s case as to the identity of the “suspect vessel” required a close calculating and comparing of times, coordinates, and distances, a process that might well have been shortcircuited by injection of the incriminating aspect of the evidence as to flashing lights. We conclude that in view of the context in which the fourth question was asked, the significance of the issue raised by that question to the outcome of tbe case, the response that was given, and the context in which the response was given, there is a reasonable possibility that the error at issue influenced the jury in reaching its verdicts in this case. Therefore, the verdicts cannot stand.
B.
Evidentiary Issues
Several of the evidentiary issues raised by the defendants are likely to recur in the event of a retrial and therefore we will address those issues. First, the defendants contest the admission of testimony from law enforcement officers who, not having been qualified as experts, offered opinion testimony at trial based on their experience.
Specifically, the defendants object to the testimony of John Alpers and Raul Rivera suggesting that items dropped from an airplane and not otherwise gathered together would have been further apart than the objects appearing on Rivera’s FLIR behind the suspect vessel; to U.S. customs agent Roberto Escobar’s testimony suggesting that the reason a wrong number might be painted on the rear of a vessel would be to hinder law enforcement authorities from identifying its owner; and to Escobar’s characterization of the coordinates of the airdrop as being “pretty close” to those found on the back of the business card found in Rivera-Santiago’s wallet.
In
United States v. Paiva,
892 F.2d 148 (1st Cir.1989), we noted that the “modem trend favors the admission of opinion testimony [from lay witnesses], provided it is well founded on personal knowledge and susceptible to cross-examination.”
Id.
at 157 (permitting drag user to express opinion that substance she found was cocaine). We further explained that “the individual experience and knowledge of a lay witness may establish his or her competence, without qualification as an expert, to express an opinion on a particular subject outside the realm of common knowledge.”
Id.; accord United States v. VonWillie,
59 F.3d 922, 929 (9th Cir.1995) (based on experience, police officer could testify as lay witness that it was common for drag traffickers to use weapons to protect drags; opinion was helpful to determination of whether defendant was involved in drag trafficking).
After reviewing the record, we find no error in the admission of the opinions of Rivera and Alpers concerning the proximity of the bales to Rivera-Santiago’s vessel. Both witnesses -testified that they had extensive experience as air interdiction officers with the U.S. Customs Service, and were competent to testify as to the behavior of objects dropped from an airplane. Further, the testimony was helpful both to the jury’s understanding of the likelihood that two objects dropped from an airplane would be close together after the airdrop and to the jury’s resolution of the question of the defendants’ involvement in the airdrop.
Nor do we find error in the admission of Escobar’s statement concerning the numbers painted on the back of Rivera-Santiago’s vessel. Escobar testified at trial that, in addition to other law enforcement experience, he had worked for a smuggling unit of the U.S. Marshals Service for two years, during which time he had been involved in approximately fifteen missions. As such, he was entitled to draw on his law enforcement experience in conveying opinion testimony to the jury. Further, his testimony concerning the significance of the incorrect registration number was helpful to the jury, which might not have been aware of the existence of a central registration system for sea vessels similar to that for automobiles.
However, we do find error in the trial judge’s decision to permit Escobar to testify that the coordinates found on the back of the business card were “pretty close” to the coordinates of the airdrop site. Immediately pri- or to offering this conclusion, Escobar testified that he did not understand what some of the numbers on the business card meant and acknowledged that he was “not an expert in coordinates.” In addition, his testimony that the coordinates were “pretty close” represented only his characterization of the distance between the two points. As presented, the witness’s testimony lacked an appropriate foundation and his conclusion was of little aid to the jury in understanding the evidence.
The defendants also challenge as unfairly prejudicial and without foundation the testimony of a U.S. drug enforcement agent, who, not having been qualified as an expert, stated on direct examination that the street value of the cocaine found at sea may have exceeded $18 million. However, we have recently stated that “[t]here is little dispute that such information may aid in proving intent to distribute.”
United States v. Rivera,
68 F.3d 5, 8 (1st Cir.1995),
cert. denied,
— U.S. —, 116 S.Ct. 970, 133. L.Ed.2d 891 (1996). In addition, in this case the evidence was relevant to bolster the government’s claim that a smuggler would not drop valuable property into the water unless a specific target was in the area. Finally, as we noted in
Rivera,
“DEA agents are especially qualified, and need not be certified as experts, to testify about street value.”
Id. We
see no error in the admission of this testimony.
C.
Sentencing Issues
Since the court has determined that the defendants’ convictions must be vacated, there is no need to address the sentencing issues raised on appeal.
III.
Conclusion
The defendants’ convictions are
vacated.
The eases are
remanded
for a new trial.