People v. LaValle (3 N.Y.3d 88, 2004)

OPINION OF THE COURT

G.B. Smith, J.

A Suffolk County jury found defendant Stephen S. LaValle guilty of first degree murder in the course of and in furtherance of first degree rape (Penal Law § 125.27 [1] [a] [vii]) and sentenced him to death. His direct appeal comes to us from Supreme Court pursuant to our unique jurisdiction in death penalty cases (NY Const, art VI, § 3 [b]; CPL 450.70 [1]). For the reasons that follow, we uphold the conviction, but vacate the death sentence, and remit for resentencing in accordance with CPL 470.30 (5) (c). In vacating the death sentence, we conclude that the jury deadlock instruction prescribed in CPL 400.27 (10) is unconstitutional under article I, § 6 of the State Constitution. We further conclude that this defect in the existing statute can only be cured by a new deadlock instruction from the Legislature.

I. FACTS

Around 12:30 pm. on Saturday, May 31, 1997, the body of Cynthia Quinn was found in the woods near Mills Road in Yaphank, a village in Suffolk County. Her neck, chest, back and arms were covered with 73 puncture wounds made with a screwdriver-like instrument. She had been raped. She also had a broken rib, bruises on her arms and abrasions on many parts of her body.

About six hours earlier, she had left her home for her customary morning jog. Several Yaphank residents saw her running along her route. By 7:30 a.m., her husband Brian Quinn grew concerned that she had not returned home and began searching for her. He had expected her to return by 7:00 a.m. because, as a self-employed carpenter, he was scheduled to report to a work site at 7:30 a.m. He also knew that Cynthia, a high school track coach, had a track meet that same morning. Initially, Brian searched the surrounding area in his car, with his two young children, and then enlisted the help of fellow volunteer firefighters. Eventually, several units of the police department joined the search. The body was discovered by two volunteer firefighters.

Earlier that same morning, Monique Sturm was attacked and robbed by a man who bumped her car on a road in Port Jefferson, which is about eight miles from where the body was found. She managed to escape through the passenger door when the man forced himself through the door on the driver’s side. In the process, she bit his finger. In her call to 911, placed at 5:57 a.m., Sturm provided a description of the assailant and his car. Sturm’s wallet was found around 11:40 a.m., not far from the murder scene.

Police investigators connected the two events, inferring that the same person was behind both. According to pretrial testimony, several weeks earlier a pedestrian was accosted by the driver of a car similar to that of defendant. Defendant was a suspect in that incident, and Sturm’s description of the assailant and his car matched the description that woman gave. The police learned that in 1986 defendant was convicted of sexually assaulting a female driver he had bumped with his car and that he was on parole as the result of a burglary conviction. The police also learned that defendant’s car was similar to a car seen near the crime scene about 6:30 a.m. that Saturday. That witness identified defendant’s car as the car he had seen. Sturm, however, could not identify defendant’s car as the one her assailant had been driving and could not definitively identify defendant in a photo array.

Two days after the murder, on June 2, 1997, the police arrested defendant in connection with the robbery after he was told to report to his parole officer. Defendant’s right index finger had a cut. Defendant waived his Miranda rights and was interrogated. According to the trial testimony of the interrogating officer, defendant initially denied any involvement with Sturm. Eventually, he admitted that his car hit Sturm’s car. Defendant told the police that the night prior to the encounter, he and his family had gone to dinner around 7:30 em. at a local restaurant. After dinner, he went to a comedy club, arriving at around 8:30 em. An hour later, he went to a pub named Harrigan’s, where he stayed for a few hours with his friend Phil Anderson and a man named Brett. After dropping off Phil at his home, defendant told the police, he and Brett went to a keg party, where they stayed for a few hours. At around 5:45 a.m., he dropped off a man named Tom near a 7-Eleven store. As defendant drove off, he turned a corner and he hit a car. When he pulled over, a female got out of the other car and began yelling at him and attacked him with her pocketbook. He took the pocketbook from her, and threw it over a fence. Defendant claimed that he tried to calm her down by pushing her into the car, but that she hit him and kicked him. Defendant admitted to the police that he got inside her car, and she jumped out the passenger door and ran to a nearby house yelling for help. Defendant said he returned to his car and drove away. Defendant prepared a sketch of where the incident took place and gave the police permission to search his car and take samples of his hair, blood and saliva.

Defendant eventually confessed to the murder of Cynthia Quinn. He stated that while driving home, he had stopped on the road to urinate. A woman jogging on the road began yelling that he was a bum and should use a restroom. He became angry because people had been yelling at him his whole life, and walked toward her. Defendant told the police that the woman backed into the woods and began waving a long thin piece of metal, like a screwdriver, at him. Defendant could see that she was scared. Defendant grabbed the piece of metal away from her and started stabbing her. When she fell down, he raped her. Then, he started stabbing her again. According to defendant, he later sat down on a nearby log and cried. Defendant told the police that he ran back to his car and drove toward his house, discarding the weapon along the way. When he got home, he threw his clothes into a hamper. Defendant drew sketches of the murder scene and the weapon for the police.

On June 5, 1997, a grand jury indicted defendant for murder in the first degree (Penal Law § 125.27 [1] [a] [vii]—intentional murder in the course of committing or attempting to commit and in furtherance of rape in the first degree, or in the course of and furtherance of immediate flight after committing or attempting to commit rape in the first degree); three counts of murder in the second degree (Penal Law § 125.25 [1], [2], [3]—intentional murder, depraved indifference murder and felony murder), and robbery in the first degree (Penal Law § 160.15 [2]). He was arraigned on June 9, 1997. On August 12, 1997, the People served notice of intent to seek the death penalty (CPL 250.40).

Opening statements commenced on June 8, 1999, and the trial lasted 17 days. The jury considered 41 prosecution witnesses and 180 exhibits, including evidence showing that defendant’s DNA matched samples collected from the victim’s body and clothing. Defendant did not present any witnesses. Two days after summations, the jury found defendant guilty of first degree murder and one count of second degree murder (intentional murder) and found him not guilty of the robbery charge. After the jury’s verdict, the trial court, pursuant to CPL 300.30 (4) and 300.40 (3) (b), dismissed the guilty verdict for murder in the second degree in light of the first degree murder verdict. The penalty phase began on August 3, 1999 and concluded on August 6 when the jury rendered a verdict that defendant be sentenced to death.

Throughout the proceedings below, defendant filed a flurry of motions, most of which the courts denied, and objected to numerous rulings. On appeal, defendant raises a number of issues. We conclude that defendant’s guilt was established beyond a reasonable doubt, and that the verdict of guilt was not against the weight of the evidence. We consider first issues arising during the guilt phase and then turn to the penalty phase. We do so with the awareness both that death is different and that CPL 470.30 confers upon this Court unique powers of review (see People v Harris, 98 NY2d 452, 474 [2002]).

II. GUILT PHASE

A. JURY SELECTION

Defendant argues that the trial court erred in failing to grant challenges for cause against jurors Nos. 16, 43, 79, 107 and 124, because they were not impartial and held views that substantially impaired their ability to consider a life sentence. None of these jurors sat on the case. The defense exhausted its peremptory strikes, and defendant argues for relief pursuant to CPL 270.20 (2), which states

“An erroneous ruling by the court denying a challenge for cause by the defendant does not constitute reversible error unless the defendant has exhausted his peremptory challenges at the time or, if he has not, he peremptorily challenges such prospective juror and his peremptory challenges are exhausted before the selection of the jury is complete.”

We disagree with defendant that the trial court erroneously failed to grant the challenges for cause.

Under CPL 270.20 (1) (b), a prospective juror who “has a state of mind that is likely to preclude him from rendering an impartial verdict based upon the evidence adduced at the trial” must be excused for cause. “[A] prospective juror whose statements raise a serious doubt regarding the ability to be impartial must be excused unless the juror states unequivocally on the record that he or she can be fair and impartial” (People v Chambers, 97 NY2d 417, 419 [2002]).

Additionally, in a capital case, a prospective juror must also be excused for cause if he or she “entertains such conscientious opinions either against or in favor of such punishment as to preclude such juror from rendering an impartial verdict or from properly exercising the discretion conferred upon such juror by law in the determination of a sentence” (CPL 270.20 [1] [f]). As this Court held in People v Harris, “Where jurors express conscientious views concerning the death penalty yet still make clear that they are able to follow their oaths to act impartially, they cannot be excluded for cause from participating on the jury” (98 NY2d 452, 484 [2002]; see also People v Cahill, 2 NY3d 14, 47 [2003]).

Defendant argues that juror No. 16 should have been excused because he had already decided defendant was guilty, he was sympathetic and felt a common bond with the victim, he believed that police officers are more credible than other witnesses, and he was biased in favor of the death penalty in first degree murder cases. In moving to dismiss juror No. 16 at trial, defense counsel argued only that juror No. 16’s answers during voir dire were totally different from those on his written questionnaire and that he had a preconceived notion about the case and felt sympathy for the victim.

During voir dire, juror No. 16 expressed doubts about his initial impression, and recognized that “the system is supposed to be fair, everyone gets a fair trial, and I believe in that.” When questioned about his views about confessions, he again stated that while he initially thought defendant was guilty, he since realized that “we have to come into the courtroom with a blank slate and listen to evidence from both sides and make a decision.” Time after time, when asked to respond to answers from his questionnaire, juror No. 16 gave answers evidencing impartiality, such as, “We have to come into the courtroom and see what the facts are and make a decision based on that,” and “I try to be fair ... I don’t always end up by being fair, but I do try to be fair ... If I’m selected, I will do the very best job I can.” Defendant’s argument that juror No. 16 would find police officers more credible than other witnesses is unpreserved. In any event, juror No. 16 unequivocally indicated that he would follow the court’s instruction not to give police testimony more weight than that of other witnesses. Juror No. 16, like the deceased, was a runner. He stated that it would not affect him, and that he would base his decision on the evidence presented at trial. Additionally, this juror assured the trial court of his ability to be impartial (see People v Johnson, 94 NY2d 600, 613-614 [2000]).

On the subject of the death penalty, juror No. 16 indicated that he would consider whatever mitigation was required by law and that he had no preconceived notion of how defendant should be punished. He therefore did not hold a conscientious belief that would prevent or substantially impair his ability to exercise discretion during the penalty phase (see Harris, 98 NY2d at 484).

B. THE SELF-REPRESENTATION CLAIM

On the same day that jury selection began, February 16, 1999, defendant requested new attorneys, complaining that he was dissatisfied with his current counsel. He complained that his two lawyers “spend no time with each other” or with him, and “want to attack this case in two different ways.” Defendant said that he wished to “get this over and get this going as quick as” possible, but that “there’s so many things left undone.” His lead attorney, Robert Gottlieb, characterized the attorney-client relationship as broken. The court denied the application, urging defendant to cooperate with his lawyers. Defendant again complained that he wished to “be able to go over my case with my lawyers and have questions answered, and there’s so many things unanswered.”

On May 18, 1999, at defendant’s request, defendant and his attorneys met ex parte with the court. The source of the conflict was crystallized. Gottlieb pointed to a severe disagreement over trial strategy. Counsel’s strategy involved making certain “concessions [with respect to culpability] that would lead to credibility all down the line.” Defendant, counsel complained, was uncooperative, even ordering his family not to cooperate with a private investigator. Defendant stated that problems over strategy arose once the prosecutor offered life without parole if defendant pleaded guilty. His lawyers and his family were urging him to accept the plea. Defendant explained that his strategy centered on denying any involvement with the murder and rape of Cynthia Quinn. As counsel put it, defendant wished to “fight everything tooth and nail to the end until it’s determined.”

After some discussion with the court in which defendant expressed his dissatisfaction with his lawyers as well as ambivalence as to whether he wished new lawyers or wanted the court to instruct his lawyers to abide by his strategy, defendant stated, “Your Honor, if you are telling me that I have to respect and listen to my lawyers [’] views on how to attack this case, I would have to disagree with you. I would ask that you would dismiss my lawyers and if I could represent myself.” The court responded, “That to me would be absolutely insane.”

Defendant remained steadfast in his view that he could not follow any strategy other than one that proclaimed his innocence, adding, “The only thing I see and that’s my last option is to represent myself, not that I want to, I don’t know [anything] about the law, but at least I have a chance to prove my innocence.” The court again responded, “[I]f you are asking me to dismiss your lawyers, I can’t do that in good conscience. If you are asking me to allow you to represent yourself pro se at this point, I can’t do that in good conscience because it would be absolutely disastrous.” The court continued, “If I allowed you to represent yourself, everything I said [about protecting your rights] would be out the window ... I just want you to think about that.” Defendant responded that if the court was going “to deny me [the right] to represent myself, maybe take into consideration appointing two new lawyers.” The court stated that it could not do that either. Explaining at length why self-representation would be a disaster, the court urged him “to think about what you have said.” The court concluded the discussion with the statement, “Let the record reflect, he does have a smile on his face.”

On May 24, 1999, a week before trial, Gottlieb moved to withdraw because there was “a total breakdown” in his relationship with defendant. Gottlieb made, clear that he could not follow defendant’s trial strategy. Associate counsel, Martin Efman, stated that he was willing to try the case as defendant wished, and defendant confirmed that he could proceed with Efman. The court granted Gottlieb’s withdrawal motion, elevated Efman to lead counsel, and approved Charles Von Schmidt as associate counsel.

While the right to counsel has emerged as a cherished and valued right, it has not displaced the constitutional right of self-representation. The request to represent oneself must be invoked clearly and unequivocally (Faretta v California, 422 US 806 [1975]; People v Davis, 49 NY2d 114 [1979]). We have held that to ensure convicted defendants not “pervert the system by subsequently claiming a denial of their pro se right, the pro se request must be clearly and unconditionally presented to the trial court” (People v McIntyre, 36 NY2d 10, 17 [1974]). Before allowing a defendant to proceed pro se, the court must conduct a searching inquiry to ensure that the waiver of the right to appointed counsel is “unequivocal, voluntary and intelligent” (People v Smith, 92 NY2d 516, 520 [1998]). Denial of the right of self-representation is not subject to harmless error analysis (McKaskle v Wiggins, 465 US 168, 177 n 8 [1984]).

We conclude that defendant’s request to represent himself was not clearly and unequivocally presented and, therefore, the court did not err in acting as it did. Defendant gave the impression that he was not committed to self-representation. Rather, he was understandably hesitant considering the daunting task of undertaking self-representation in a capital case. After all, even lawyers who defend capital cases must receive special training. Initially, defendant made the conditional statement that if the court forced him to go along with the strategy of his attorneys, he would ask to represent himself. Defendant did not assertively state that he wanted to represent himself. Defendant then stated that he saw self-representation as his last option, though he did not want to represent himself because he did not know anything about the law. When defendant mentioned self-representation for the last time, he again couched it as a hypothetical, adding the request for new lawyers as an alternative. The court urged defendant to think about the perils of self-representation and did not make a definitive ruling. Defendant’s statements do not reflect a definitive commitment to self-representation.

In contrast to this case, in Williams v Bartlett, a case that defendant relies upon, the defendant asserted, “I will represent myself’ (44 F3d 95, 97 [1994]) after being warned of the dangers of self-representation. The defendant had represented himself before the grand jury. On another occasion, defendant stated, “it’s . . . my intention[ ] now to go pro se. Before I wanted to have an attorney, but I can’t afford a private attorney. That’s why I’m going pro se” (id. at 98). The United States Court of Appeals for the Second Circuit found that on each occasion, defendant’s “statements show a ‘purposeful choice reflecting an unequivocal intent to forego the assistance of counsel’ ” (id. at 100, quoting United States v Tompkins, 623 F2d 824, 827-828 [2d Cir 1980]). In this case, defendant wavered, and when warned at length about the perils of self-representation, and asked to think about his request, he smiled and said nothing more.

While conditioning a request for new attorneys with a request for self-representation does not necessarily make the latter request equivocal, it is clear that in this case defendant raised the specter of self-representation as a means of procuring the dismissal of Gottlieb. Defendant and Gottlieb were intractably at odds over strategy. Defendant had sought to have Gottlieb removed on two occasions prior to the May 18th meeting. When Gottlieb announced in a meeting with the court and the other attorneys on May 24th that he wished to be relieved, the court later asked defendant for his view, and he stated, “I’m ready to proceed . . . with . . . Martin Efman. Me and Mr. Gottlieb cannot get along . . . Our whole trial strategies are different . . . Me and Mr. Efman are willing to go to trial the way I would like to go to trial.” The court later relieved Gottlieb. After the dismissal, defendant never said a word about wanting to represent himself. The issue of self-representation was closed, with defendant satisfied with Gottlieb’s dismissal, the elevation of Efman to lead counsel and the appointment of a new associate counsel.

C. WITHHOLDING OF EVIDENCE—BRADY MATERIAL

More than two months after the commencement of jury selection and more than two years after the murder, on April 29, 1999, the police obtained a written statement from John Doe, a friend of defendant. Doe stated that “[a]bout two years ago,” defendant had asked him to “hang out.” They went to another house, belonging to Richard and Maria, where the three men “partied and drank beers” in the basement. Doe and defendant then drove to a bar, although neither drank there. They then drove around, and “had a couple of more beers.” Because he was tired, Doe called it a night. When defendant drove Doe home, it was still dark. Several days later, Doe heard that defendant had been arrested.

Richard Roe provided two statements. In the first, dated May 5, 1999, he stated that he met defendant during the Memorial Day weekend (May 24-26) of 1997. A couple of days later, on a Friday, he, defendant and John Doe went into the basement of his house where they “partied” for a few hours. Defendant drank two beers and seemed “docile.” According to Roe, defendant and Doe stayed for a few hours and then left.

The second statement, dated May 19, 1999, differed significantly from the first. Roe maintained that Doe and defendant arrived at Roe’s house around 9:30 or 10:30 em. Defendant had two beers, and they all smoked crack. They then left the house to purchase more crack from a drug dealer. The dealer wanted defendant “to fuck someone up and LaValle said he’d do it.” Roe allegedly told Doe to “get LaValle back, to leave.” Roe claimed he “was afraid what might happen.” According to this version, the three went back to the basement, and continued to smoke crack until about 1:30 to 2:00 a.m. when Roe told defendant and Doe that they had to go. Roe also claimed that he became nervous when defendant told him that his then-15-year-old daughter was pretty, and that guys would pursue her. The next day, a friend told Roe that defendant was a “dirt bag.” A couple of days later, he learned about the murder.

On June 2, 1999, about a week before trial, the prosecutor informed the defense that the police had obtained statements from a few witnesses who might have been with defendant the night before the murder. On July 1, defense counsel requested copies of the statements. Counsel referred in particular to the Doe statement, which he believed might have contained information beneficial to defendant. The court told the People to submit any statements falling under Brady v Maryland (373 US 83 [1963]) for in camera review. The prosecutor consented, adding that he had provided the defense with “a general outline” of the Doe statement. Defense counsel expressed the concern that “the People were trying to hold [any statements] for rebuttal if we called [the witnesses] in the first instance” and that it was “important now for us to see what’s in there or for the court, at least, to make a determination.” Defense counsel added that the defense had attempted to interview Doe, but that he was “not very giving of information” and that “the most telling information might be in the statement he gave to the police.” In his written submission, defense counsel argued that despite reasonable diligence, the defense could not ascertain the contents of the statement.

In a written opinion, the court ruled that the statements were not Brady material since defendant knew of the contents of the statements and also knew that Roe and Doe had provided the statements. The court also ruled that during the penalty phase, the prosecutor would satisfy his Brady obligation by providing defendant with the identity of witnesses who may be a source of relevant mitigation evidence. After the close of evidence in the guilt phase, but before summations, defense counsel again requested the statements, arguing Roe and Doe were not forthcoming “about certain areas which might incriminate them.” Defense counsel raised the issue again after summation, arguing that the statements might be relevant as a mitigator, but the court adhered to its prior ruling.

On July 30, defense counsel renewed his motion for the disclosure of the Doe and Roe statements for use in the penalty phase. In his attached affirmation, counsel stated that prior to trial, the prosecutor had informed the defense of the existence of written statements by two witnesses. Counsel believed that the statements contained information that “defendant, together with these individuals, was under the influence of alcohol and narcotics for many hours immediately preceding the incident which is the subject of the indictment and First Degree Murder conviction.” This information was relevant to the circumstances of the case and as a mitigator. Counsel argued that the statements were admissible at the penalty phase as reliable hearsay and statements made against penal interest. In the other affirmation, defense counsel Charles Von Schmidt detailed his unsuccessful personal efforts to speak with Doe.

In Strickler v Greene, the Supreme Court summarized the “essential components of a Brady violation” (527 US 263, 280 [1999]) as follows: “The evidence at issue must be favorable to the accused, either because it is exculpatory, or because it is impeaching; that evidence must have been suppressed by the State, either willfully or inadvertently; and prejudice must have ensued” (id. at 281-282). Brady held “that the suppression by the prosecution of evidence favorable to an accused upon request violates due process where the evidence is material either to guilt or to punishment, irrespective of the good faith or bad faith of the prosecution” (373 US at 87).

We do not doubt that the statements contained some measure of favorable evidence. We conclude, however, that they were not suppressed by the prosecution and that, accordingly, there was no Brady violation in this case. Evidence is not suppressed where the defendant “knew of, or should reasonably have known of, the evidence and its exculpatory nature” (People v Doshi, 93 NY2d 499, 506 [1999]). Defendant knew or should have known that he drank alcohol and took drugs with Roe and Doe. Moreover, defense counsel, in an affirmation seeking the statements, admitted that the defense believed that the statements dealt with use of alcohol and drugs the night before the murder. The defense knew the identities and addresses of Roe and Doe. Possession of the statements would not have revealed any essential information that the defense did not already know (see Allen v Lee, 366 F3d 319, 325 [4th Cir 2004] [where defendant argued that the government withheld jail records indicating that he received antiwithdrawal medication, no Brady claim arose since defendant “had personal knowledge of any medication he might have received”]). Finally, we note that the prosecution submitted the statements to the court which decided that the prosecution had satisfied its Brady obligation.

Defendant argues that this case is indistinguishable from Leka v Portuondo (257 F3d 89 [2d Cir 2001]), where the Second Circuit determined that the government violated its Brady obligation. We disagree. Leka was a murder case in which the prosecutor’s evidence consisted of the testimony of a number of eyewitnesses. Early in the case, the prosecutor informed the defense that one of the witnesses was a police officer who would identify the defendant. Three business days before the trial, however, the prosecutor told defense counsel that the officer would not identify the defendant, but did not say anything about what the officer saw. Because the defense used a ruse to try to contact the officer, the court granted the prosecutor’s request to foreclose access to the officer.

The Second Circuit held that the witness’s favorable testimony was suppressed, finding that “the prosecutor never made specific disclosure of what [the witness] had seen. There is no doubt that the prosecutor had that information from the beginning of the case” (257 F3d at 100). The court rejected the government’s argument that disclosure of the witness’s name, and (presumably) the address, satisfied its Brady obligation, noting that the disclosure was “too little, too late” (id.). The court reasoned that “[t]he limited Brady material disclosed to Leka could have led to specific exculpatory information only if the defense undertook further investigation” (id. at 101). The court also noted that the defense used the ruse because the government failed to be forthcoming about what the officer knew. In addition, it was too much to expect defense counsel to seek an order vacating or modifying the court’s order precluding contact with the officer on the eve of tried when counsel did not know whether the witness would provide useful information.

There are significant differences between Leka and this case. The defense in Leka knew nothing about what the witness knew. Here, the defense knew the essential information contained in the statements. The disclosure here was not too little, nor was it too late. The defense had a week prior to the start of the trial in which to subpoena Roe and Doe.

Defendant also relies on Boss v Pierce (263 F3d 734 [7th Cir 2001]). In Boss, the state relied on the testimony of one eyewitness to establish that two brothers, along with others, killed and robbed the victim. Defendants called four witnesses, one of whom was his girlfriend who provided an alibi. The girlfriend’s sister partially corroborated her testimony. On the last day of the two-day trial, the state gave defendants an investigative report containing information unknown to defense counsel (id. at 737-738), namely that the sister told investigators that the government’s witness had implicated himself in the killing and robbery, and only implicated the brothers to get them in trouble. The United States Court of Appeals for the Seventh Circuit rejected the state’s argument that since the sister was defendants’ witness, the defense had access to any information she possessed. The court held that defendants had no reason to suspect that the sister would know the information in the report. Again, in this case, the defense knew the essential information contained in the report.

Another case defendant relies on, United States v Severdija (790 F2d 1556 [11th Cir 1986]), in which the United States Court of Appeals for the Eleventh Circuit found a Brady violation, is closer to this case. There, the captain of a ship was convicted of conspiracy to possess with intent to distribute marijuana found on his ship. Days before his arrest, the defendant told a member of the Coast Guard who boarded his ship that he had been hired to pick up a tow, that he did not know his crew very well and that he did not trust them. Defendant also suggested that the Coast Guard remain in the area because the boat would be carrying drugs. The member of the Coast Guard recorded defendant’s statements. At the trial, defendant testified that he informed the Coast Guard about the marijuana, which had been placed on his boat over his objection.

While acknowledging that the defendant must have been aware of the statement he made, the court found that the evidence at issue was the recordation of the statement, not the statement itself. The statement could have corroborated defendant’s defense. The court noted that had defendant testified about the statement, without corroborating evidence, jurors would likely have perceived it as a self-serving statement. It is significant, however, that the court reached this result after finding that “neither the identity of the [Coast Guard officer who recorded the statement] nor the contents of his report was disclosed to the defense until after the jury had returned its verdict” (id. at 1559). That was not the case here. Moreover, in this case, defendant did not present any testimony during the guilt phase that the statements would have corroborated. Indeed, defendant claimed he was not the person who committed the crime. The second Roe statement would have further tarnished defendant before the jury, by portraying defendant as a crack cocaine user, a voyeur of Roe’s young daughter, a “dirt bag,” and a person willing to assault another at the behest of a drug dealer.

D. “VICTIM-IMPACT” EVIDENCE

During the guilt phase the prosecutor opened with the statement that on the day of the murder, “Suffolk County became a lesser place. . . . On that day, we, as a community, lost Cindy Quinn to a brutal act of random predatory violence as she jogged near her home in Yaphank. Cindy was someone we all know, a mother, a wife, a daughter, a sister, a friend, a teacher, a coach.” The prosecutor would urge the jury “to show [defendant] the same compassion he showed to [the mother of two young children].”

On June 14, defendant moved to preclude testimony of Brian Quinn, Cynthia’s husband, arguing that the testimony would inflame the jurors. Defense counsel offered to “stipulate to everything, her death and body being in that location.” The prosecution refused. The trial court allowed the testimony based on a timeline test, demarcating events prior to the murder as permissible, and those after as impermissible.

Quinn testified that he was a self-employed carpenter and that he built the family home with his own hands. On the day of the murder, he was awakened by the cries of his two-year-old son. Quinn identified a ruby earring his wife was wearing when she was found. He also provided other personal information regarding his wife—that she was a high school teacher with a Master’s degree, she coached the school’s track team, she had two living sisters, her parents were alive, and he and the deceased also had a four-year-old daughter. Every morning she would wake up near dawn to jog and would sometimes run with the children in a jogging stroller.

The defense objected to much of the testimony, characterizing it as victim-impact evidence. Defendant argues here that much of Quinn’s testimony constitutes an improper victim-impact statement through which the prosecutor advanced the theme of lost promise.

While it is only natural for jurors to react with horror at the brutality of a violent murder, jurors have an obligation to decide the issues in the case “in a judicial temper. Appeals to sympathy or prejudice can but be harmful” (People v Caruso, 246 NY 437, 443 [1927]). In Caruso, the prosecution, over objection, called the widow of the deceased who testified that they had been married for 18 months, had a six-month-old child, and that the deceased would sit on the baby’s crib and sing to her. The widow’s testimony, we concluded, constituted reversible error because it “had no materiality upon the issues before the jury” and was an “ ‘unseemly and unsafe’ appeal to prejudice” (id. at 444). In People v Miller (6 NY2d 152 [1959]), the prosecutor elicited from the deceased’s brother that the deceased had a wife and seven children. We again found, “There could be no purpose to this fine of testimony but to conjure up in the minds of the jurors undue prejudice against the defendant” (id. at 157).

In People v Harris, where the type of testimony of victims’ families was “indistinguishable from that in Miller and Caruso,” we stated, “Although family information about a victim is an important aspect of the victim’s life, generally, it has no bearing on defendant’s guilt or innocence” (98 NY2d at 491). There we found that, to the extent the claims were preserved, the error was harmless in light of the overwhelming evidence of guilt.

Here, unlike Caruso and Miller, much of the testimony was material because it explained the sequence of events, from the husband’s discovery that his wife was missing to learning that his wife’s body had been found later in the day. To be sure, there was testimony that was irrelevant to the sequence of events, such as that Cynthia had a Master’s degree, but such testimony was minimal and harmless where the proof was overwhelming and there was no significant probability that, but for this error, the jury would have acquitted defendant (see People v Crimmins, 36 NY2d 230, 237 [1975]).

E. GUILT PHASE SUMMATION

Defense counsel’s summation reflected somewhat the initial conflict that permeated defendant’s relationship with his lawyers. Counsel argued that defendant was not the person who committed the murder. Alternatively, he argued that the person who did it had demons inside, and suddenly snapped at the slightest provocation. Counsel maintained that the circumstances surrounding the killing evinced lack of control. He also urged that “there’s some greater system in the world that determines when and how we die,” in an effort to shift blame away from defendant.

The prosecutor began by noting that “mere words are very inadequate to convey the horror of what occurred to Cindy Quinn that morning two years ago.” He added symbolically, “I know you have visited those woods on that Saturday attempting to see and feel what atrocity went on. I remind you now what the testimony of Lynn Weyant. . . My God. What sort of scream came from those woods to prompt that type of reaction? What sort of torture caused that scream? Now you know. You’ve seen it. You’ve seen what he did to Cindy Quinn.”

The prosecutor referred to Sturm as “quivering and shuddering” on the witness stand as she confronted her attacker, and contrasted her good fortune to the brutal death of Cynthia Quinn. Sturm “was lucky enough to go on with her life and now she has a beautiful little baby boy.” Cynthia’s two children, on the other hand, “have no mommy,” Brian has no wife, and her family has lost a daughter and a sister. The prosecutor told the jurors, “We can only pray that Cindy thought to take the time to pat them on the head or kiss them before she went out that morning.”

In conclusion, the prosecutor told the jurors that for 17 days they had “sat in the presence of evil.” And unlike defendant, Cynthia Quinn was buried 25 months earlier without the opportunity to present a defense. The prosecutor asked the jurors to look defendant in the eye and tell him, “[Ylou’re a murderer, you’re a rapist, and you’re guilty, and you shall be held responsible.”

After the prosecutor’s summation, defense counsel moved for a mistrial, arguing that the prosecutor sought to inflame the jury. The court denied the motion.

On appeal, defendant argues that the prosecutor committed a number of improprieties during the guilt phase summation: inviting jurors to relive the victim’s suffering by expressing that words could not describe what she must have experienced; referring to Sturm’s good fortune and Quinn’s cruel fate; lamenting the pain of those Cynthia Quinn left behind; degrading defendant by characterizing him as a killer whale; arguing that defendant received the benefit of due process while Cynthia Quinn did not; and urging the community to avenge, and thereby heal.

Defendant’s arguments are similar to those raised in People v Harris and People v Cahill. In Harris, the defendant claimed “that the prosecutor invited the jury to punish defendant for exercising his rights to a trial and not to testify, that the prosecutor misstated the record, referred to facts not in evidence, and misused evidence, and that the prosecutor denigrated defendant’s defense and inflamed the jury” (98 NY2d at 491 n 18). We held that the claims were unpreserved, and, had we reviewed them, we would have concluded that the statements did not “exceed the broad bounds of rhetorical comment permissible in closing argument” (id. [citation omitted]). We came to the same conclusion in Cahill, where the defendant argued that “the People made improper remarks throughout the trial, alluding to, among other things, the victim’s beauty and courage and defendant’s remorselessness” (2 NY3d at 72 n 45).

As in Harris and Cahill, we conclude that defendant did not properly preserve the claims he makes on appeal. The motion for a mistrial was insufficient to preserve the arguments. Nevertheless, reviewing the merits of the claims, we conclude that while several of the prosecutor’s remarks were improper, reversal is not warranted (see Crimmins, 36 NY2d at 237).

III. PENALTY PHASE

We next address the constitutionality of the “deadlock instruction” the court delivered to the jury prior to its deliberation on the appropriate sentence for defendant. Pursuant to CPL 400.27 (10), the court instructed the jurors on their duty to decide whether defendant should be sentenced to death or to life without parole. Either choice had to be unanimous. The court further instructed the jurors, as required by statute, that if they failed to agree, the court would sentence defendant to life imprisonment with parole eligibility after serving a minimum of 20 to 25 years. Although defendant did not object when the instruction was given, he had sought a ruling prior to trial that CPL 400.27 (10) was unconstitutional under the Federal and State Constitutions, both on its face and as applied. His motion preserved the argument for appeal.

Like some other states with death penalty statutes, New York recognized that jurors should know the consequences of a deadlock (see Berberich, Note, Jury Instructions Regarding Deadlock in Capital Sentencing, 29 Hofstra L Rev 1301, 1324 [Summer 2001]). However, New York’s deadlock provision is unique in that the sentence required after a deadlock is less severe than the sentences the jury is allowed to consider. No other death penalty scheme in the country requires judges to instruct jurors that if they cannot unanimously agree between two choices, the judge will sentence defendant to a third, more lenient, choice.

Studies have found that jurors tend to “grossly underestimate how long capital murderers not sentenced to death usually stay in prison” (Bowers and Steiner, Death by Default: An Empirical Demonstration of False and Forced Choices in Capital Sentencing, 77 Tex L Rev 605, 648 [Feb. 1999]). Jurors’ beliefs with respect to the actual number of years a defendant will serve in prison are compelling and can even be decisive. As the study concluded, the “sooner jurors think a defendant will be released from prison, the more likely they are to vote for death and the more likely they are to see the defendant as dangerous” (id. at 703). A study of South Carolina jurors who served in capital cases “confirm[ed] that jurors’ deliberations emphasize dangerousness and that misguided fears of early release generate death sentences” (Eisenberg and Wells, Deadly Confusion: Juror Instructions in Capital Cases, 79 Cornell L Rev 1, 4 [Nov. 1993]; see also Garvey, Aggravation and Mitigation in Capital Cases: What Do Jurors Think?, 98 Colum L Rev 1538, 1560 [Oct. 1998] [finding that “(f)uture dangerousness appears to be one of the primary determinants of capital-sentencing outcomes”]). Thus, jurors might impose the death penalty on a defendant whom they believed did not deserve it simply because they fear that the defendant would not serve a life sentence. These studies provide the best available insight into jury behavior.

The New York Legislature determined that defendants convicted of capital first degree murder should either be sentenced to death or the most severe alternative sentence, life imprisonment without the possibility of parole. What, then, is the consequence of telling the jury that it may not impose a sentence of life with parole eligibility after 20 to 25 years, but that the court will impose that sentence if the jury cannot agree? The deadlock instruction interjects the fear that if jurors do not reach unanimity, the defendant may be paroled in 20 years and pose a threat to society in the future. Yet, in New York a defendant’s future dangerousness is not a statutory aggravator the jury may consider.

By interjecting future dangerousness, the deadlock instruction gives rise to an unconstitutionally palpable risk that one or more jurors who cannot bear the thought that a defendant may walk the streets again after serving 20 to 25 years will join jurors favoring death in order to avoid the deadlock sentence. Commentators have been harshly critical of New York’s deadlock instruction. For jurors who are inclined toward life without parole, the choice is between death and life with parole, a Hobson’s choice in light of the jurors’ likely concerns over defendant’s future dangerousness. The choice of death results not through “a comparison of views, and by arguments among the jurors themselves,” but through fear and coercion (Jones v United States, 527 US 373, 382 [1999], quoting Allen v United States, 164 US 492, 501 [1896]). New York’s statute is unique in its coercive effect.

The case closest in point to the one before us because of the instruction given to the jury is Morris v Woodford (273 F3d 826 [9th Cir 2001], cert denied 537 US 941 [2002]). There, the Court of Appeals for the Ninth Circuit held that because of a mistake, there was a reasonable likelihood that one or more jurors interpreted the court’s instructions to mean that if the jury could not agree between death and life without parole, the defendant would be sentenced to life with parole. In other words, the coercive instruction mistakenly given in Morris v Woodford is the same instruction actually required here by CPL 400.27 (10). Remanding the case for a new penalty-phase trial, the court concluded that the instructions “would suggest to any holdout juror that, if he or she did not join the majority of the other jurors, then Petitioner would be eligible for parole. That suggestion is, of course, incorrect, and its coercive potential is obvious; in effect, it would place such a juror in the apparent position of choosing between death and life with parole” (273 F3d at 841).

Here, the trial court reasoned that it was more likely that jurors favoring life without parole would not switch to death since lack of unanimity would mean that the defendant would remain incarcerated for at least 20 to 25 years, an option preferable to death. But under the existing statute, life with parole was not one of the options available to the jury. New York’s Legislature did not intend for jurors to choose life with parole. A juror who purposely chooses that option would fail to abide by the court’s instruction to decide between death or life without parole. In any event, a juror who has found defendant guilty of a capital crime, and has heard weeks of arguments and a summation reviling the defendant and detailing the pain he has caused, is more likely to choose death than risk the prospect of defendant ever harming anyone in society again (see Deadly Confusion, supra, 79 Cornell L Rev at 12 [“The data suggest that the sentencing phase of a capital trial commences with a substantial bias in favor of death”]).

It is of course possible that one or more jurors who favor death might be coerced into choosing life without parole rather than have defendant be sentenced to life with parole. That, however, does not cure the coercive effect of the deadlock instruction before us. The coercive effect is not relieved by recognizing that some jurors may be coerced in the opposite direction.

Because death is qualitatively different, there is a “corresponding difference in the need for reliability in the determination that death is the appropriate punishment in a specific case” (Woodson v North Carolina, 428 US 280, 305 [1976]). Whether a juror chooses death or life without the possibility of parole, the choice is driven by the fear that a deadlock may result in the eventual release of the defendant. Under New York’s deadlock instruction the choice is not, as it should be, the result of a reasoned understanding that it was the appropriate one. We hold today that the deadlock instruction required by CPL 400.27 (10) is unconstitutional under the State Constitution because of the unacceptable risk that it may result in a coercive, and thus arbitrary and unreliable, sentence.

A. LEGISLATIVE DEBATE ON THE DEADLOCK INSTRUCTION

The New York State Legislature was aware of the danger that the deadlock instruction might result in coercive verdicts. During deliberations on the death penalty statute, Senator Richard Dollinger addressed the issue of coercion. He asked, “[H]ow do you avoid the problem of a jury that is hung up on the issue of either life in prison without parole or the death penalty of putting additional pressure on the jurors, knowing that if they failed to agree they are going to face a penalty that is less than either of the two penalties that they are currently in dispute over?” (New York State Senate Debate, Mar. 6, 1995, at 1912.) Senator Dale Volker responded that he did not believe the procedure was coercive, stating that in other jurisdictions “there is ample constitutional basis” for that conclusion (id.). As Senator Volker explained, “the jury—in order to enact the death penalty, it has to be a unanimous decision. If there’s no unanimous decision, if they can’t arrive at a decision for life without parole, then the judge has the option then to make the decision to sentence the person to 25 years to life” {id. at 1912-1913).

Senator Bollinger persisted, asking whether giving the jury all three options had been considered. Senator Volker replied that such a procedure had been considered, as had the alternative of giving the judge the sentencing option of either life without parole or 20 to 25 years to life. The feeling was “if you are going to have a jury enact something as serious as the death penalty, you ought to have a unanimous jury” {id. at 1915). Therefore, “the decision was to give the jury the option for the most severe penalties and let them know that if they couldn’t choose one of those more severe penalties then the judge would enact the lesser penalty which is 20-25 years to life” {id. at 1915-1916). Senator Bollinger again asked, “But isn’t it inherently coercive to tell them that you have to do this; otherwise there is going to be another penalty imposed?” {Id. at 1916.) Senator Volker replied that there would be a problem if the jury was not informed as to the consequences of a hung jury. He thought that “that would be a much more serious constitutional problem” (id.).

B. GUIBING AUTHORITY

The United States Supreme Court has not ruled on an instruction by which jurors are told that if they cannot agree on a verdict, the defendant will receive a lesser sentence than the ones they must consider. To be sure, the Supreme Court has held that defendants, and capital defendants in particular, are entitled to uncoerced verdicts (see Lowenfield v Phelps, 484 US 231 [1988]). In Simmons v South Carolina (512 US 154 [1994]), the Court held that if the state makes future dangerousness an issue, and the alternative sentence to death is life without parole, the defendant should be allowed to argue that he is parole ineligible. In other words, “[t]he State may not create a false dilemma by advancing generalized arguments regarding the defendant’s future dangerousness while, at the same time, preventing the jury from learning that the defendant never will be released on parole” (id. at 171). Here, the choice of life without parole is rendered illusory by the fear that failure to unanimously agree will result in a parole eligible sentence.

Moreover, in Beck v Alabama (447 US 625 [1980]), the Supreme Court invalidated a death penalty scheme that precluded jurors from considering a noncapital lesser included offense. The Court held that the failure to give jurors the power to convict defendant of the lesser included offense “would seem inevitably to enhance the risk of an unwarranted conviction. Such a risk cannot be tolerated in a case in which the defendant’s life is at stake. As we have often stated, there is a significant constitutional difference between the death penalty and lesser punishments” (id. at 637). The Court explained that “a jury might convict a defendant of a capital offense because of its belief that he is guilty of some crime, or, given the mandatory nature of the death penalty under Alabama law, the jury might acquit because it does not think that the defendant’s crime warrants death” (California v Ramos, 463 US 992, 1007 [1983] [discussing Beck v Alabama]). The statute interjected “irrelevant considerations into the factfinding process, diverting the jury’s attention from the central issue of whether the State has satisfied its burden of proving beyond a reasonable doubt that the defendant is guilty of a capital crime” (447 US at 642).

Although Beck involved the guilt/nonguilt phase, much of its reasoning applies to this case. Here, the statute precludes the jury from considering a third option—one that many jurors might find unpalatable—and yet it instructs that the third option will be imposed upon a deadlock. It thereby improperly interjects future dangerousness, a nonstatutory aggravator, into jury deliberations. It is unclear, however, whether the Supreme Court would apply Beck’s reasoning to this case because in California v Ramos (463 US at 1009), the Court held that “the risk of an unwarranted conviction is simply not directly translatable” to the penalty phase. There, the defendant challenged the instructing of jurors that a life sentence without parole could be commuted by the Governor. The Court found that the instruction focused the jury on the individual characteristics of the defendant and his offense, was accurate, and could be rebutted by defendant.

Subsequently, the California Supreme Court held the instruction unconstitutional on state due process grounds because it was “misleading and because it invites the jury to consider speculative and impermissible factors in reaching its decision” (People v Ramos, 37 Cal 3d 136, 159, 689 P2d 430, 444 [1984]). In so holding, the California high court quoted approvingly from the New Jersey decision in State v White (27 NJ 158, 177178, 142 A2d 65, 76 [1958])—as do we:

“It is no more proper for a jury to conclude that death be the penalty because a life sentence may be commuted or the defendant paroled, th[a]n it would be for a trial judge in other criminal causes deliberately to impose an excessive sentence to frustrate the statutory scheme committing parole to another agency. That death should be inflicted when a life sentence is appropriate is an abhorrent thought.”

In concluding that the deadlock instruction is unconstitutional, we are persuaded by the rationale of those two high courts.

The People rely on Jones v United States (527 US 373 [1999]) where the defendant argued that because of the trial court’s erroneous instruction, the jury believed that the defendant would receive a sentence of less severity than life without parole upon a deadlock. To avoid the possibility of the defendant’s possible release, jurors who favored life without the possibility of parole could have changed their votes to death. The defendant had requested the trial court to instruct the jury that in the event of a deadlock, the sentence would have been life without the possibility of parole. In a divided opinion, the Supreme Court initially held that “the Eighth Amendment does not require that the jurors be instructed as to the consequences of their failure to agree” (id. at 381).The Court reasoned that the jury was not “affirmatively misled” by the trial court’s refusal to give the proposed instruction, which “has no bearing on the jury’s role in the sentencing process” (id. at 381, 382). Although the proposed instruction would have been accurate, the Court found that it might undermine the strong governmental interest in securing unanimity. The Court declined to follow state cases that have required jurors to be informed as to the consequences of a deadlock.

Reviewing for plain error, the Court also held that the trial court did not mislead jurors into believing that the deadlock sentence would be of less severity. But even if the court had misled the jurors, the defendant could not “show the confusion necessarily worked to his detriment” since it was “just as likely that jurors, loath to recommend a lesser sentence, would have compromised on a sentence of life imprisonment as on a death sentence” (527 US at 394).

According to the People, the principle underlying Jones and Ramos is that as long as the information conveyed to the jury is accurate, it need not be balanced. Absent from the Court’s discussion in Jones is any mention about the heightened need for reliability in death penalty cases. Instead, the Court focused on the importance of securing jury unanimity, and the fact that the jury was not affirmatively misled. New York’s deadlock instruction gives jurors accurate information, and tilts toward unanimity. The risk of coercion, however, calls into question the reliability of the jury’s determination. Under our case law, the goal of jury unanimity may not undermine sentencing reliability, particularly where the sentence is death.

C. NEW YORK’S PRECEDENT ON COERCED VERDICTS

We have held that a coerced verdict “ought not to be allowed to stand in any case, and least of all, in one involving a human life” (People v Sheldon, 156 NY 268, 285 [1898]). Since Sheldon, we have consistently affirmed this principle. In People v Faber (199 NY 256, 259 [1910]), we held that “[t]he verdict of a juror should be free and untrammeled. In arriving at a verdict the judge presiding at the court must not attempt to coerce or compel the jury to agree upon a particular verdict, or any verdict” (see also People v Dawkins, 82 NY2d 226 [1993]; Matter of Plummer v Rothwax, 63 NY2d 243 [1984]; People v Carter, 40 NY2d 933 [1976]).

Generally, the source of the coercion has been the trial court. In well-meaning, though misguided, attempts to get juries to reach a decision after all the travails of a trial, courts have given coercive instructions. Although shocking to imagine, there was a time when jurors would be deprived of food and rest until they reached a verdict. In rejecting the practice, Chancellor Kent reasoned that a verdict reached under such conditions is “founded not on temperate discussion and clear conviction, but on strength of body” and “does not . . . stand with conscience, but is altogether repugnant to a sense of humanity and justice” (People v Olcott, 2 Johns Cas 301, 309 [1801], quoted in People v Sheldon, 156 NY at 276). In People v Sheldon, we expressed sympathy with the trial court’s desire to see the case end with a verdict, but concluded that the court “fell into error, and, as a result, very likely coerced some members of the jury into an agreement with their fellow-members against their own personal convictions” (id. at 282-283). The error in Sheldon was keeping the jurors together in a cramped room for about 85 hours without beds or cots, and giving the impression that they would remain confined for longer if they did not reach a verdict.

Other jury instructions we have invalidated have not been as bluntly coercive as in Olcott or Sheldon. In People v Faber, the trial court instructed that a juror “should join with his co-jurors, and should make in some respects their opinion his own” (199 NY at 258). We held that “[t]he trial judge, in his appar