OPINION OF THE COURT
Graffeo, J.
The primary issue in each of these buy-and-bust cases is whether the trial court properly closed the courtroom to the general public during the testimony of undercover officers. We conclude that the limited closures comported with Sixth Amendment public trial principles, but a new trial is required in one case based on an erroneous jury charge on the agency defense.
I.
Echevarria
Defendant Alex Echevarria was charged with criminal sale of a controlled substance in the third degree and criminal sale of a controlled substance in or near school grounds for selling three bags of crack cocaine to undercover officer number 0076 while undercover officer number 0064 (the “ghost”) observed from a distance. The buy occurred on St. Nicholas Avenue near West 155th Street in northern Manhattan. The court held a pretrial hearing pursuant to People v Hinton (31 NY2d 71 [1972]) to determine whether public access should be restricted during the testimony of the two undercover officers on the ground that closure was necessary to protect their safety.
At the hearing, officer number 0076 testified that he remained active in the area of St. Nicholas Avenue, returning to the vicinity 15 to 20 times since defendant’s arrest. He had a number of pending cases and “lost subjects”—sellers who had evaded apprehension—from St. Nicholas Avenue, and the officer expected to return there “[a]ny time soon.” He took precautions to conceal his identity during his frequent courthouse visits and had been threatened by a suspect on one occasion. Officer number 0064 gave similar testimony regarding his undercover activities around St. Nicholas Avenue and West 155th Street. [8]*8He too had been threatened, once with a knife, and a suspect had hit him in the head with a hard object a few weeks before the Hinton hearing.
Over defendant’s objection, the court concluded that the courtroom should be closed to the general public during the testimony of both officers. The court reasoned that the officers had made a particularized showing that open court testimony would jeopardize their safety and ongoing investigations. The court noted that it would separately consider opening the proceedings to defendant’s family members should they wish to attend.
At trial, officer number 0076 testified that defendant offered to take the officer’s $40 to a nearby building to obtain crack cocaine. When the officer was hesitant about parting with his money, defendant gave him his wallet to hold as collateral. Defendant returned a short time later and handed the officer three bags of crack cocaine. Officer number 0064 observed the transaction. Two arresting officers testified that, shortly after the sale, they searched defendant and found a bag of crack cocaine in his pocket. A criminalist stated that laboratory testing established that the three packets defendant handed to the officer contained narcotics and an expert witness for the People described the nature of typical street narcotics transactions. Defendant took the stand and testified that he was a drug addict who agreed to procure the drugs for the undercover officer because the officer told him that he could keep one bag of crack cocaine.
The trial was open to the public save for the testimony of the two undercover officers. The court gave the jury a charge on the agency doctrine. The jury rejected defendant’s agency defense and convicted him as charged. He was later sentenced to two concurrent prison terms of 10 years followed by three years of postrelease supervision. The Appellate Division affirmed (89 AD3d 545 [1st Dept 2011]), and a Judge of this Court granted defendant leave to appeal (18 NY3d 957 [2012]).
Moss
Defendant Andrew Moss was charged with criminal sale of a controlled substance in the third degree for selling 10 bags of crack cocaine to undercover officer number 2454. Undercover officer number 5986 acted as a ghost during the operation. The sale took place on the corner of West 135th Street and Broadway in northern Manhattan. Before trial, the court held a Hinton [9]*9hearing to decide whether to close the courtroom during the testimony of the undercover officers as a safety measure.
Officer number 5986 worked in an area that included the 30th precinct, which encompassed West 135th Street and Broadway, and testified that he had made 15 to 20 narcotics purchases in the immediate vicinity subsequent to defendant’s arrest—one as recently as a week before the hearing. He had a number of pending cases and lost subjects from that area. The officer had also been threatened and searched by suspects on prior occasions and took steps to maintain his anonymity in the courthouse.
Officer number 2454 primarily worked in the 33rd precinct— immediately adjacent to the 30th precinct—but remained active in the 30th precinct. He testified that he had made an additional 30 to 40 buys in the immediate vicinity of West 135th Street and Broadway after defendant’s arrest and had 10 pending cases arising from purchases in that area. He had previously been threatened with guns, knives and scissors, and one suspect had placed a contract on his life. He had also been patted down 50 times by suspects and had encountered them in and around the courthouse on prior occasions. The officer therefore took steps to preserve his cover when he entered a courthouse to testify, including using a private entrance normally reserved for judges.
At the conclusion of the Hinton hearing, defense counsel objected to the courtroom closure. As an alternative, defense counsel suggested that the court could post a court officer at the door to “screen” visitors. In response to defense counsel’s comment that courtroom closure was unnecessary because it was unlikely that anyone would attend, the court noted that probation violators frequently sat in the audience awaiting their proceedings. The court determined that it would close the courtroom during the testimony of both officers because the People had sufficiently demonstrated that open court testimony would jeopardize their safety. The court made an exception for family members, however, and defendant’s mother attended the full trial.
At trial, undercover officers, numbers 2454 and 5986, testified about the drug transaction and identified defendant as the seller. The arresting officer, a criminalist and an expert also testified for the People in open court. The jury convicted defendant as charged, and he was subsequently sentenced to 10 years’ imprisonment with three years of postrelease supervi[10]*10sion. The Appellate Division affirmed (89 AD3d 600 [1st Dept 2011]), and a Judge of this Court granted defendant leave to appeal (18 NY3d 960 [2012]).
Johnson
Defendant Martin Johnson was charged with criminal sale of a controlled substance in the third degree and criminal sale of a controlled substance in or near school grounds for selling a small quantity of crack cocaine to undercover officer number 206 while undercover officer number 0014, the ghost, watched from a short distance away. The transaction took place on East 132nd Street and Park Avenue in East Harlem. The court conducted a pretrial Hinton hearing to ascertain whether it was necessary to close the courtroom during the officers’ testimony to protect their safety.
At the hearing, officer number 206 stated that almost all of his 10 open cases involved buys “within the vicinity where this case took place.” He had been in the area of defendant’s arrest as recently as two weeks before the hearing and expected to return there for future undercover work. In addition, he had 20 subjects who were still at large. He took precautions when entering the courthouse and had been threatened by drug dealers in the past. The officer had also been “struck up side the head” and had an object thrown at him. Officer number 0014 similarly testified that he had a number of open cases emanating from “the same vicinity as where this case occurred.” He remained active in that area after defendant’s arrest, participating in 50 additional buys during that time period, and expected to be assigned there for future investigations. He also took precautions to shield his identity when he visited courthouses and had been threatened with weapons multiple times while working undercover. On one occasion, he was forced to dive behind a car when a suspect opened fire on him.
After the hearing, defense counsel objected to the courtroom closure. The prosecutor responded that there was “no lesser option” than closure to protect the officers’ safety. The court granted the People’s application “to the limited extent” of ordering closure during the testimony of the undercover officers, excepting defendant’s family members.
At trial, officer number 206 testified that defendant took $20 from him, entered a building and returned with crack cocaine. Officer number 0014 observed the transaction. The arresting officer and a criminalist also testified in open court. The court [11]*11gave a jury charge on the agency defense, which the jury rejected when it convicted defendant as charged. He was sentenced to concurrent three-year prison terms to be followed by two years’ postrelease supervision. The Appellate Division affirmed (88 AD3d 503 [1st Dept 2011]), and a Judge of this Court granted defendant leave to appeal (19 NY3d 1103 [2012]).
II.
All three defendants maintain that they are entitled to a new trial because they were deprived of their constitutional right to a public trial due to the exclusion of the general public during the testimony of the undercover officers. Although a criminal defendant’s Sixth Amendment right to an open trial is fundamental, it is not absolute (see People v Martin, 16 NY3d 607, 611 [2011]).1 Rather, trial judges have “inherent discretionary power to exclude members of the public from the courtroom” (People v Ramos, 90 NY2d 490, 497 [1997] [internal quotation marks and citations omitted]). Nevertheless, there remains a presumption of openness. As a result, the right to a public trial “may yield to other rights or interests in rare circumstances only, and the balance of interests must be struck with special care” (id. [internal quotation marks and citations omitted]). A violation of this right is not subject to harmless error analysis and mandates reversal.
The United States Supreme Court has explained that a courtroom closure must satisfy a four-part standard to comport with the requirements of the Sixth Amendment:
“[T]he party seeking to close the hearing must advance an overriding interest that is likely to be prejudiced, the closure must be no broader than necessary to protect that interest, the trial court must consider reasonable alternatives to closing the proceeding, and it must make findings adequate to support the closure” (Waller v Georgia, 467 US 39, 48 [1984]).
Only the first prong (overriding interest) and third prong (reasonable alternatives) are at issue in these cases.
A.
Moss and Johnson, but not Echevarria, argue initially that the People failed to demonstrate a sufficient likelihood of prejudice [12]*12to an overriding interest to justify closing the courtroom during the testimony of the undercover officers under the first prong of the test. In particular, Moss asserts that undercover officer number 2454’s safety concerns were geographically unfounded because he was generally assigned to the 33rd precinct whereas Moss’s arrest occurred in the 30th precinct. Johnson similarly contends that the Hinton hearing testimony regarding the geographic area in which both undercover officers continued to work was too vague or generalized to support closure under prong one. We disagree.
We have previously had occasion to analyze the first Waller prong in the specific context of officer safety in undercover buy- and-bust operations (see People v Jones, 96 NY2d 213 [2001]; Ramos, 90 NY2d 490; People v Ayala, 90 NY2d 490 [1997] [companion case to Ramos]; People v Martinez, 82 NY2d 436 [1993]; People v Pearson, 82 NY2d 436 [1993] [companion case to Martinez]; see also People v Jones, 47 NY2d 409 [1979]; Hinton, 31 NY2d 71).
The safety of law enforcement officers “unquestionably” may constitute an overriding interest (Ramos, 90 NY2d at 498). Yet, “the mere possibility that this safety interest might be compromised by open-court testimony does not justify abridgement of a defendant’s constitutional right to a public trial” (id.). The United States Supreme Court has made plain that the proponent of closure must demonstrate a “substantial probability” that the identified interest will be prejudiced by an open courtroom (Press-Enterprise Co. v Superior Court of Cal., County of Riverside, 478 US 1, 14 [1986] [Press-Enterprise II]). Hence, a nexus or “specific link must be made between the officer’s safety concerns and open-court testimony in the particular buy-and-bust case” (Ramos, 90 NY2d at 498; see also Jones, 96 NY2d at 217).
In Martinez, for example, we held that the trial judge improperly closed the courtroom where the undercover officer testified that he continued to work “in the Bronx area” and generally feared for his safety (82 NY2d at 443). We concluded that the first prong was not met by this “perfunctory showing,” emphasizing that the officer’s reference to the Bronx area—a borough consisting of more than 41 square miles and 1.2 million people—without greater specificity was insufficient (id.). We also noted that the officer did not indicate whether he had ever received a threat, nor did he identify any targets of investigation who could potentially show up in the courtroom.
[13]*13On the other hand, the requisite nexus was established in Ramos. In that case, both undercover officers continued to operate in the 104th precinct—the precinct where defendant was arrested—and expected to resume undercover operations there shortly. Moreover, the officers had observed former “buy” subjects at the courthouse on prior occasions and had several cases pending before the court (90 NY2d at 499). They each took precautions to conceal their identities when entering the courthouse and one gave specific testimony about a gunpoint threat he had received when a former subject discovered he was a police officer. Under these circumstances, we upheld the closure of the courtroom during the officers’ testimony.
Similarly, in Ayala, although a “closer question,” we concluded that there was a specific link between the undercover officer’s safety concerns and his open-court testimony (90 NY2d at 499). There, the officer identified three precincts in which he was recently active and expected to return imminently. These included the precinct where he bought drugs from the defendant and the precinct in which the courthouse was situated. Though the officer could not pinpoint the area of future operations with any greater precision, we determined that the testimony was sufficiently specific, explaining that the officer’s reference to “certain precincts was akin to identifying the particular neighborhoods in which he continued to be active” (id. at 500). We also noted that the officer had a number of ongoing investigations in the three precincts; he had been previously threatened in front of the courthouse by a defendant in another case; and he used a private entrance to the courthouse. “On balance,” we determined that the trial court did not abuse its discretion in finding the first prong satisfied (id.).
Moss’s case falls comfortably within the ambit of Ramos and Ayala. Although defendant notes that undercover officer number 2454 was mainly assigned to a precinct other than the one where defendant was arrested, the officer made clear that he continued to operate in the precinct of the arrest as well.2 Indeed, the officer made dozens of buys in the immediate vicinity following defendant’s arrest. He had a number of open cases from the area of the arrest and stated that he could be assigned to that particular area at any time. Based on this track record, the inference was strong that the officer would return to the [14]*14specific location of defendant’s arrest for future undercover work. The officer also referred to numerous gun and knife threats he had received; the previous encounters he had with suspects in the courthouse; and the steps he took to protect his identity when entering courthouses. Taken together, the record amply supports the trial court’s determination that a specific link existed between the officer’s safety and his open-court testimony.
The overriding interest prong was also met in the Johnson case. Both undercover officers stated that they remained active in the “vicinity” where defendant’s case “took place.” Indeed, undercover officer number 0014 engaged in dozens of narcotics purchases in the “area of this case” subsequent to defendant’s arrest. Both officers also expected to return to the neighborhood for future undercover operations. Although defendant now claims that the officers’ references to the “area” or “vicinity” of his arrest—without specifically referring to street names— were too vague, he did not voice this concern at the Hinton hearing. In any event, the obvious implication of the testimony was that the area or vicinity to which the officers were referring meant the localized area of defendant’s arrest. The officers’ testimony was therefore sufficiently specific under our precedents. Finally, both officers had been threatened or physically attacked with weapons, including guns, and took care to conceal their identities when at courthouses. Under these circumstances, the trial court did not err in finding a nexus between the officers’ safety and their testimony in an open courtroom.
Having concluded that prong one of the Waller test was satisfied, we now turn to defendants’ prong three claims.
B.
Echevarria, Moss and Johnson all contend that their constitutional right to a public trial was violated because the trial judge in each case failed to comply with the third prong of Waller— the requirement that courts consider reasonable alternatives to closure. They assert that prong three places an obligation on trial courts to expressly consider and comment on alternatives to closure on the record, and that the failure to do so automatically necessitates reversal and a new trial. They rely heavily on Presley v Georgia (558 US 209 [2010]), which held that trial courts are to consider alternatives to closure even when they are not suggested by the parties. Defendants claim that the trial courts below should have discussed the possibility of having the [15]*15undercover officers testify from behind a partition or posting a court officer at the door to “screen” entrants to the courtroom.
In response, the People acknowledge that trial courts are required to evaluate alternatives to closure pursuant to prong three. But the People submit that, under Ramos, trial courts need not take the additional step of explaining those alternatives on the record in those cases where “it can be implied that the trial court, in ordering closure, determined that no lesser alternative would protect the articulated interest” (Ramos, 90 NY2d at 504). The People posit that Presley did not overrule this aspect of Ramos and, therefore, we should reaffirm Ramos under principles of stare decisis. Although the dissent disagrees, we believe that the People have the better argument.
In Ramos, we addressed and squarely rejected prong three arguments identical to those defendants raise before us now. The trial court in Ramos (and Ayala, the companion case) closed the courtroom during the testimony of undercover officers to protect their safety but did not explicitly discuss alternatives short of closure on the record. We considered whether the trial courts’ failure to vocalize alternatives under the circumstances was compatible with the requirements of the third Waller prong. At the outset of our analysis in Ramos, we agreed with the defendants that prong three requires “trial courts, before excluding the public, to consider whether something short of complete closure would protect the ‘overriding interest’ at stake” (Ramos, 90 NY2d at 502-503). Indeed, we noted that Waller itself mandated that trial judges “must consider reasonable alternatives to closing the proceeding” (id. at 503, quoting Waller, 467 US at 48). But we also observed that Waller “does not hold that the trial court must explicitly consider alternatives on the record” to satisfy its prong three obligation (id.). To resolve this question—the very question before us on the present appeal—we surveyed the United States Supreme Court cases that had found prong three violations.
We started with Waller, where the record plainly did not justify the trial court’s closure of the entire seven-day suppression hearing. Only a small portion of the hearing—totaling 21fe hours—involved sensitive information that the government sought to shield from the public, and no showing was made that the remaining six-plus days of the hearing should have been closed. In light of this facial record deficiency, we stated that the Waller trial court’s “overly broad closure order could not be upheld without some explanation for its failure to limit closure to [16]*16the sensitive portions of the proceeding” (id.). Put differently, because it was apparent on the face of the record that the trial court in Waller did not consider reasonable alternatives—most notably limiting closure to the 2V2-hour period—the United States Supreme Court found a prong three violation (see Waller, 467 US at 48-49 [“The court did not consider alternatives to immediate closure of the entire hearing: directing the government to provide more detail about its need for closure . . . and closing only those parts of the hearing that jeopardized the interests advanced”]).
Next, we looked at Press-Enterprise Co. v Superior Court of Cal., Riverside Cty. (464 US 501 [1984] [Press-Enterprise I]). In that case, in an effort to protect sensitive information relayed by potential jurors and encourage juror candor, the trial court closed all but three days of a six-week voir dire in a rape and murder trial at the government’s behest. We noted that, as in Waller, the proponent of closure “failed to adduce adequate record facts to support the conclusion that the entire voir dire proceeding would involve sensitive juror questioning” (Ramos, 90 NY2d at 503). Despite this glaring record deficiency, the trial court in Press-Enterprise I “made no effort to limit closure to those portions of the proceeding comprising sensitive subject matter” (id.). The trial court later refused to release a transcript of the closed proceeding, even though it acknowledged that the questioning was “of little moment” (Press-Enterprise I, 464 US at 504). Under these circumstances, the United States Supreme Court concluded that there was a prong three “failure to consider alternatives to closure and to total suppression of the transcript” (id. at 513).
Finally, we evaluated Press-Enterprise II, a case in which the trial court in a high profile murder trial granted the defendant’s unopposed motion to close the entirety of a preliminary hearing to the public.3 Thereafter, the trial court refused to release a transcript of the 41-day-long closed proceeding. We stated that the United States Supreme Court was clearly “concerned about the excessive breadth of the closure” in that case (Ramos, 90 NY2d at 503). The Supreme Court held that the trial court violated prong three by failing to consider [17]*17“alternatives short of complete closure,” noting that “closure of an entire 41-day proceeding would rarely be warranted” (Press-Enterprise II, 478 US at 14-15).
In reviewing these United States Supreme Court decisions in Ramos, we found that the records in all three “lacked minimally sufficient facts to establish that the portions of the proceedings subject to closure would even implicate the interests sought to be protected” (Ramos, 90 NY2d at 503). We explained that the need to more narrowly design closure in those cases was apparent from the record because only those parts of the proceedings that prejudiced an overriding interest could be closed. In contrast, the trial court in Ramos (and Ayala) narrowly tailored the courtroom closures to the portions of the proceedings for which overriding safety interests were implicated, i.e., during the testimony of the undercover officers. Ultimately, we held that where the record in a buy-and-bust case “makes no mention of alternatives but is otherwise sufficient to establish the need to close the particular proceeding ... it can be implied that the trial court, in ordering closure, determined that no lesser alternative would protect the articulated interest” (id. at 503-504 [emphasis added]). In other words, we rejected the precise argument defendants now press—that the trial court must always explicitly explain that it has considered but rejected alternative measures to comply with prong three.
Nevertheless, defendants and the dissent urge that Ramos is no longer good law in light of Presley. During jury selection in Presley, the trial court observed a lone spectator in the courtroom, petitioner’s uncle. The court instructed the uncle to leave, explaining that the entire courtroom was needed to seat prospective jurors and that there was no space for the public. When defense counsel requested an “accommodation,” the trial court refused, stating that each row was needed for the 42 prospective jurors and that a family member “cannot sit and intermingle with members of the jury panel” (Presley, 558 US at 210). After petitioner’s conviction, he moved for a new trial based on the exclusion of the public from the voir dire. In support of his motion, he adduced evidence establishing that 14 prospective jurors could have fit in the jury box and the remaining 28 could have been seated on one side of the courtroom, leaving sufficient space on the other side for the public. The trial court denied the motion and the Supreme Court of Georgia affirmed, reasoning that trial courts need not consider alternatives to closure unless the opposing party proactively offers a suggestion.
[18]*18The United States Supreme Court reversed in Presley and ordered a new trial, holding that Waller itself made “clear” that, under the third prong, “trial courts are required to consider alternatives to closure even when they are not offered by the parties” (558 US at 214). The trial court’s exclusion of the public during the voir dire was deemed improper because the third prong was not satisfied on the record:
“Trial courts are obligated to take every reasonable measure to accommodate public attendance at criminal trials. Nothing in the record shows that the trial court could not have accommodated the public at [petitioner’s] trial. Without knowing the precise circumstances, some possibilities include reserving one or more rows for the public; dividing the jury venire panel to reduce courtroom congestion; or instructing prospective jurors not to engage or interact with audience members” (558 US at 215).
The core teaching of Presley is that prong three obliges trial courts to consider reasonable alternatives to closure, even where the parties themselves do not bring those alternatives to the attention of the court. To the extent Ramos is read to suggest otherwise, Presley certainly controls. But contrary to the dissent’s view, Presley did not break new ground,4 and it nowhere states that it is incumbent on trial courts, regardless of the circumstances, to engage in a verbal on-the-record review of all potential alternatives before opting for a limited closure. Nor do we read Presley as calling into question our holding in Ramos that the record in a given buy-and-bust case may support the conclusion that the trial judge impliedly considered alternatives before closing the courtroom. Indeed, in stark contrast to Ramos, the record in Presley made clear that the trial judge’s exclusion of the public (and petitioner’s family member in particular) was unwarranted and that the space constraints could have easily been remedied using less intrusive measures. Hence, we adhere to that portion of Ramos that held that the absence of explicit discussions regarding alternatives is not fatal under prong three where the record in a buy-and-bust case otherwise suffices to establish the need to close a particular portion of the proceeding. Under those circumstances, “it can be implied that the trial court, in ordering closure, determined that no lesser alternative [19]*19would protect the articulated interest” (Ramos, 90 NY2d at 504).
Applying Ramos here, we perceive no prong three violation in the cases now before us. As in Ramos, the trial court in each instance held a Hinton hearing and made a particularized finding that requiring the undercover officers to testify in open court would create a genuine risk to their physical safety. The trial courts limited the closures to the portions of the proceedings directly implicating the overriding interest—the undercover officers’ safety—by ordering the courtrooms closed only for the duration of the officers’ testimony. Moreover, even those portions were only partially closed in the Moss and Johnson proceedings because the trial courts in those cases made an exception for defendants’ family members to attend. Similarly, at the Echevarria trial the court stated that it would separately consider opening the closed portion to family members if they requested access. On the record before us, it is fair to imply that the trial courts concluded that no lesser alternative would have adequately protected the officers’ safety and, therefore, the courts discharged their prong three duty to consider reasonable alternatives.5
In sum, we conclude that defendants were not deprived of their Sixth Amendment rights to a public trial by the trial courts’ narrowly-tailored closures. But we caution that courtroom closure is not available just for the asking in buy-and-bust cases. As we emphasized in Ramos, a criminal defendant’s Sixth Amendment rights “must not be lightly cast aside simply because the People claim that an undercover officer’s safety or effectiveness is at risk, and trial courts must vigilantly ensure that Waller’s demanding first prong is satisfied before closing a courtroom” (id. at 506). Those sentiments are no less true today.
III.
Apart from their public trial claims, Echevarria and Johnson separately assert that they are entitled to a new trial because the trial court in each of their cases gave an erroneous jury [20]*20charge on the agency defense. They both contend that the courts’ failure to enumerate all six agency factors listed in the pattern Criminal Jury Instructions (CJI) constitutes reversible error. Echevarria further maintains that the trial court erred by instructing the jury that a finding of no prior relationship between the undercover officer and defendant would “negate” the agency defense. The People counter that the jury charge given in both cases, read as a whole, properly conveyed the agency concept and that, in any event, neither defendant was entitled to a charge on the facts presented.
Under the agency doctrine, “one who acts solely as an agent for the buyer of narcotics cannot be convicted of the crime of selling those narcotics, notwithstanding that the act of transferring drugs to another falls within the statutory definition of ‘sell’ ” (People v Ortiz, 76 NY2d 446, 448-449 [1990]). It is not a complete defense, however, in that it acknowledges wrongdoing (possession) on the part of the defendant (see People v Watson, 20 NY3d 182, 190 [2012]). A defendant is entitled to a jury charge on the agency defense where “there is some reasonable view of the evidence that the defendant acted as a mere instrumentality of the buyer” (People v Roche, 45 NY2d 78, 86 [1978]).
In assessing whether the defendant acted solely to accommodate the buyer, the jury may consider a number of factors, including:
“the nature and extent of the relationship between the defendant and the buyer, whether it was the buyer or the defendant who suggested the purchase, whether the defendant has had other drug dealings with this or other buyers or sellers and, of course, whether the defendant profited, or stood to profit, from the transaction” (Watson, 20 NY3d at 186 [internal quotation marks and citation omitted]).
But we have also made clear that jurors must rely on their “own common sense and experience” in resolving the agency issue, and that the jury “should be able to perceive the reality of the situation . . . without minutely detailed . . . instruction from the court” (People v Lam Lek Chong, 45 NY2d 64, 75 [1978] [internal quotation marks and citation omitted]). At bottom, there is no “legal formula” for determining whether a defendant was the buyer’s agent (id.).
Applying these principles, the jury charge in Johnson’s case was proper. The court appropriately instructed the jury [21]*21regarding the underlying theory of the agency defense and proceeded to list the first five factors set forth in the CJI charge.6 The jury was certainly capable of gathering from the language of the charge “the correct rules which should be applied in arriving at [a] decision” (People v Umali, 10 NY3d 417, 427 [2008] [internal quotation marks and citations omitted]). Hence, there was no error.
We reach a different conclusion in the Echevarria case. There, the trial court limited its charge to two of the factors found in the CJI, both of which were unfavorable to defendant. Moreover, the court instructed the jury that the lack of a prior relationship between defendant and the undercover officer would negative the agency defense. But our precedents and the CJI charge specify that the nature and extent of a relationship is but one factor (albeit an important one) for a jury to consider in assessing the agency defense. Hence, the imbalanced charge did not convey the proper analytical framework and was therefore error.
Contrary to the People’s position, the error was not harmless. On this record, there was a reasonable view of the evidence that defendant was acting as the undercover officer’s agent. Defendant testified that he was a drug addict (a user not a seller) who offered to help out someone whom he believed to be a fellow addict. He denied initiating the transaction and had never been convicted of selling drugs. There was also evidence that, prior to [22]*22the transaction, the drugs were exclusively within the control of a drug dealer in the nearby building. Although defendant testified that the officer gave him a bag of crack for his assistance, the receipt of an incidental benefit does not in itself negate an agency defense (see Lam Lek Chong, 45 NY2d at 75-76). We conclude that Echevarria is entitled to a new trial.
Accordingly, in Echevarria, the order of the Appellate Division should be reversed and a new trial ordered; in Moss and Johnson, the orders of the Appellate Division should be affirmed.