Holland v. Donnelly

216 F. Supp. 2d 227, 2002 U.S. Dist. LEXIS 8629, 2002 WL 999307

• Court: District Court, S.D. New York
• Decided: May 14, 2002
• Docket: 01 Civ. 2292(GEL)
• Status: Published

Opinion

OPINION AND ORDER

LYNCH, District Judge.

In this petition for habeas corpus, petitioner Darryl Holland challenges his conviction for first-degree murder and other crimes in the County Court, Westchester County (Hon. Mark C. Dillon, Judge), and resulting sentence of life imprisonment without parole. The evidence at trial, taken in the light most favorable to the prosecution, established that Holland murdered Philip Kaufman on August 30, 1996. The prosecution’s theory, amply supported by the evidence, was that Holland met Kaufman, a long-time acquaintance, in the parking lot of the supermarket where Kaufman worked, and asked him for a ride. During the ride, Holland stabbed and strangled Kaufman, stole his money, and smashed his head with a rock

Holland raises four constitutional challenges to his conviction, arguing that (1) incriminating statements he made while in police custody were involuntary products of psychologically coercive interrogation; (2) the statements resulted from renewed interrogation after he had invoked his right to remain silent; (3) his sentence to life imprisonment without parole is unconstitutional because it was not imposed by the jury; and (4) the trial court erroneously rejected his claim that the prosecutor had unconstitutionally challenged a juror on the basis of race. Because none of these claims have merit, the petition is denied.

BACKGROUND

I. The Crime

Since Holland does not challenge the sufficiency of the evidence to support his conviction, the facts of the offense can be summarized briefly. The evidence recounted in this section is drawn from the transcript of the trial.

Philip Kaufman left work at the supermarket at about 11:30 p.m. on Friday night, August 30, 1996. The police found Kaufman’s body at about 1:00 a.m. that same night on a nearby road, lying about ten feet from his car. His wallet was empty, although he had been paid and had cashed his paycheck earlier that day. The victim had been stabbed and he had suffered severe head injuries, apparently having been beaten by a large rock that was found on his head, covered with blood. (The medical examiner was later to testify that two of the stab wounds, as well as the head injuries, would have been sufficient to cause death; the victim had also been strangled, apparently with an electric cord found at the scene, though not mortally.) A white T-shirt stained with blood (later identified by DNA testing as Kaufman’s) was also recovered from the scene, as was a knife similar to one Holland had been seen carrying.

Holland, whose girlfriend was a co-worker of Kaufman’s, had spent the day smoking crack with a Mend. He left the friend’s home at around 9:00 p.m., wearing a white T-shirt and black shorts, as well as a blue sweatshirt borrowed from the friend, to get money to buy more drugs. A supermarket employee, who knew Holland, observed him speaking to Kaufman in the parking lot after Kaufman left work; he last saw the two men walking toward Kaufman’s car. Holland returned to his friend’s home about midnight, shirtless and with blood on his body and shorts, and gave him money to buy more crack. Asked what had happened, he said that “Philip put up a fight.”

A search of Holland’s apartment, conducted pursuant to a warrant not challenged here, recovered various items including bloodstained black shorts and sneakers. Other physical evidence also linked Holland to the murder. A fingerprint matching his was recovered from Kaufman’s ear; fibers under the victim’s fingernails matched the blue sweatshirt, which was recovered near the crime scene; similar fibers were also found on the bloody white T-shirt.

In addition to all of this evidence, the prosecution introduced a number of statements made by Holland, including a detailed confession, first made orally to the officers, and then repeated on videotape. The circumstances surrounding Holland’s confession are the subject of the first two claims in the instant petition.

II. The Statements

The following account of the events leading up to Holland’s confession are taken primarily from the thorough findings of fact made by the trial judge after an evi-dentiary hearing on Holland’s motion to suppress his statements. These findings of historical fact must be “presumed to be correct” for purposes of this petition, 28 U.S.C. § 2254(e)(1); Boyette v. Lefevre, 246 F.3d 76, 88 (2d Cir.2001), and in any event Holland makes no effort to challenge or refute them.

At about 5:00 p.m. on Sunday, September 1, police officers, having learned that Holland had been seen with Kaufman on the night of the murder, visited Holland at his apartment. In a statement whose admissibility is not questioned, Holland admitted that he was in the vicinity of the supermarket late on the night of August 30, and agreed to accompany the officers to police headquarters. In response to police questioning, conducted in an unlocked conference room and lasting about 35^0 minutes, Holland claimed that he had asked Kaufman for a ride home, which was refused, and then walked home.

The police left Holland in the interview room at about 6:05 p.m. He never asked if he could leave, and was never told by the police that he either was or was not free to go. At around 8:00 p.m., the officers left to interview Holland’s girlfriend, who contradicted his account, stating that he had never returned home that night. After that interview, the officers considered Holland a suspect. They returned to the police station at about 9:15 p.m., administered Miranda warnings, and obtained Holland’s signature on a form acknowledging and waiving his rights. The state trial judge later concluded that from this point on, Holland was in custody, as a reasonable person would no longer have felt free to leave. People v. Holland, No. 96/1266, Decision and Order at 13-14, 15 (Westchester County Court Sep. 24, 1997) .(“Decision and Order”).

The officers then confronted Holland with his girlfriend’s statement. Holland “did not verbally respond, but appeared nervous, rocked back and forth in his chair, and banged his head on the conference table.” Decision and Order at 5. The principal interrogator then went over Holland’s earlier statement with him; Holland adhered to what he had previously said. This interrogation ended by 11:00 p.m.

Holland spent the night in the conference room at the police headquarters, sleeping in his chair at least some of that time. The officers, meanwhile, spent the night preparing a search warrant application for Holland’s apartment and executing the warrant. 1 At 9:00 a.m. on the morning of September 2, the officers returned and confronted Holland with the bloody clothing. He acknowledged that the shorts and sneakers were his, but did not respond when asked to explain the blood.

At 10:45 a.m., after officers had recovered additional items from the crime scene, including the blue sweatshirt and knife, two detectives re-interviewed Holland. One detective, LaRotunda, attempted to persuade Holland to confess by referring to the penalties that might face him:

LaRotunda used two coffee cups and a cigarette lighter to draw analogies of the defendant’s options — the cigarette lighter representing the death penalty, a coffee cup representing life in prison, and the second coffee cup representing “seven to ten” years in jail. LaRotunda advised that the lenient option would allow Holland to see his son grow up. LaRotunda made no promises of guarantees to the defendant at any time.... When LaRotunda made reference to the defendant’s parents, Holland began to sob and then described, in detail, how he killed Philip Kaufman.

Decision and Order at 9. The defendant was then moved to a different room, where, beginning at 11:30, the confession was repeated on videotape. Holland was re-advised of his rights, executed a written waiver, and stated that he was willing to speak to the police. He declined coffee, stated that he was not hungry, and then repeated in chilling detail how he had stabbed and strangled Kaufman as Kaufman was driving him home, causing the car to crash into an embankment, stolen his money, and then beaten Kaufman to death when he asked to go to the hospital, in fear of being held accountable for the stabbing and robbery. He also identified as his the various items that had been recovered from the crime scene and at his apartment.

Holland was with the police from 5:00 p.m. on Sunday evening through about 12:30 Monday afternoon, a total of over 19 hours; for all but the first four of those hours he was in custody. He was offered food and drink on two occasions during the period of his detention, which he declined the first time and accepted the second. He was allowed to use the toilet whenever he asked. He was not interrogated continuously; after the two' relatively brief periods of interrogation referred to above, he was questioned again at 9:00 a.m. the following morning, and then yet again from 10:45 until he confessed shortly thereafter, a confession repeated in a videotaped interview that lasted from 11:30 to 12:05. (Decision and Order at 8-9, 11, 18.) Although he was not provided with a bed, he was not disturbed from 11:00 p.m. until 9:00 a.m. the following morning, and was observed sleeping in the conference room during at least some of that period. During the videotape interview, he declined coffee, indicated that he was not hungry, and stated that the police had been “very, very nice to me” and that he had not been coerced. (Decision and Order at 11, 18.) It is not contended that the police ever applied any physical force, or even raised their voices or browbeat Holland in any way.

III. Tidal and Sentence

Holland does not challenge the sufficiency of the evidence set forth above to sustain the jury’s guilty verdicts, nor does he raise any issue about the admission of evidence or court’s instractions to the jury. He does argue that the prosecution’s challenge of one potential juror was racially motivated, and that his sentence to life imprisonment without parole was imposed unconstitutionally. The facts relating to these claims can be described more efficiently in the portions of this opinion dealing with them.

DISCUSSION

1. Involuntary Confession

Holland’s first claim is that under the totality of the circumstances, his confession was not voluntary, due to the length of his detention, the “psychological pressures brought to bear” on him, and the “repeated references to the death penalty [and] the use of a vivid ‘death penalty analogy.’ ” (Pet.Br.12.) This claim was presented to and rejected by the state courts on the merits. People v. Holland, 268 A.D.2d 536, 703 N.Y.S.2d 57, 58 (2d Dep’t 2000), leave to appeal denied, 95 N.Y.2d 835, 713 N.Y.S.2d 142, 735 N.E.2d 422 (2000) (Wesley, J.); Resp. Exs. C, G. 2 Accordingly, it may only be the basis for relief in this court unless the state court’s conclusion was “contrary to, or involved an unreasonable application of, clearly established federal law, as determined by the Supreme Court of the United States.” 28 U.S.C. § 2254(d)(1).

There has been considerable appellate discussion of what it takes to find an “unreasonable application of’ Supreme Court precedent. The Supreme Court has helpfully advised that unreasonable means not reasonable, see Terry Williams v. Taylor, 529 U.S. 362, 409-12, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000), and the Second Circuit has in turn glossed that to mean not merely erroneous, but characterized by “[s]ome increment of incorrectness beyond error.” Francis S. v. Stone, 221 F.3d 100, 111 (2d Cir.2000); see also Fuller v. Gorczyk, 273 F.3d 212, 219 (2d Cir.2001). That increment does not need to be so great as to suggest “judicial incompetence”; it just needs to be—unreasonable.

While the governing test may be difficult to reduce to a helpful formula, however, it is not difficult to apply to this case, where Supreme Court precedent offers a well-established test, and the state courts applied it in a manifestly reasonable manner. As the Supreme Court stated in Dickerson v. United States, 530 U.S. 428, 434, 120 S.Ct. 2326, 147 L.Ed.2d 405 (2000), the correct test for determining whether a statement is voluntary or coerced under the due process clause is

“whether a defendant’s will was overborne” by the circumstances surrounding the giving of a confession. Schneckloth [v. Bustamonte], 412 U.S. [218,] 226, 93 S.Ct. 2041, 36 L.Ed.2d 854 [(1973)]. The due process test takes into consideration “the totality of all the surrounding circumstances—both the characteristics of the accused and the details of the interrogation.” Ibid.

See also Colorado v. Connelly, 479 U.S. 157, 163-64 & n. 1, 107 S.Ct. 515, 93 L.Ed.2d 473 (1986). While neither the trial court nor the Appellate Division cited specific federal cases, the trial court made perfectly plain that it understood that “psychological coercion [is a recognized] basis for suppressing statements,” and that in evaluating such a claim, “the Court must examine the totality of circumstances to determine whether statements uttered by a defendant were the product of psy chological pressure that overbore the defendant’s will.” (Decision and Order at 17.) The state courts thus applied the correct legal standard for determining the voluntariness of Holland’s confession.

The state courts’ application of the governing “totality of the circumstances” standard is reasonable by any definition. It can of course be argued, as Holland does, that an interrogation is coercive when it is conducted after lengthy custody, without food, under uncomfortable conditions, and with implicit warnings that cooperation will likely lead to reduced punishment. But the trial court carefully weighed this argument against the actual circumstances of Holland’s confinement in a careful opinion, pointing out that Holland was not subjected to lengthy interrogation (he was questioned on only four occasions, most of them quite brief); that “[t]here is no evidence that the defendant was sleep deprived, as he did appear to be sleeping in the interview room during the night”; that “[flood and drink were offered to the defendant on more than one occasion” (and that Holland in fact declined refreshment during the videotape interview at the end of the interrogation and stated he was not hungry); and that the “overall conduct of the ... police is best described by [Holland] himself who stated, on videotape, that the police were ‘very, very nice to [him].’ ” (Decision and Order at 18.) These considerations strongly support the court’s ultimate conclusion that under the totality of the circumstances, the defendant’s confession was not the involuntary product of police coercion. Even if the judge might reasonably have reached a different conclusion, the decision actually reached by the state judge, who was able to observe the defendant and evaluate the witnesses’ credibility after presiding over an extremely extensive evidentiary hearing, cannot remotely be characterized as “unreasonable.”

Holland places considerable weight on LaRotunda’s mention of the death penalty, and what he characterizes as an implicit promise of leniency in exchange for a confession. Although he recognizes the applicability of the “totality of the circumstances” standard of voluntariness (Pet. Br.18), Holland seeks comfort in a different formulation, citing dictum from the venerable case of Bram v. United States, 168 U.S. 532, 542-43, 18 S.Ct. 183, 42 L.Ed. 568 (1897) (quoting 3 H. Smith & A. Keep, Russell on Crimes and Misdemeanors 478 (6th ed. 1896)), to the effect that a confession may not be obtained by “ ‘any direct or implied promises, however slight.’ ” Bram, however, like Boyd v. United States, 116 U.S. 616, 6 S.Ct. 524, 29 L.Ed. 746 (1886), dates from an era in which the Supreme Court interpreted constitutional restrictions on federal law enforcement far more strictly than it does at present, or, for that matter, than it did in the heyday of the Warren Court. 3 Unlike the Court’s decision in Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966), which attempts to offset the coercive effects of custodial interrogation by requiring the famous warnings, dictum in Bram went far toward suggest ing that the inherent coerciveness of such interrogation might well render any custodial confession inadmissible under the Fifth Amendment. See Bram, 168 U.S. at 556, 18 S.Ct. 183 (noting “doubt in England” as to whether “the mere fact of the interrogation of a prisoner by a police officer would per se render the confession inadmissible, because of the inducement resulting from the very nature of the authority exercised by the police officer”). But these remarks do not represent present law. Indeed, in an opinion joined by Justices Brennan, Marshall and Stevens, the Supreme Court specifically disavowed the very passage relied on by Holland, ruling that “under current precedent [that dictum] does not state the standard for determining the voluntariness of a confession.” Arizona v. Fulminante, 499 U.S. 279, 285, 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991).

Nor are Holland’s claims of “repeated references to the death penalty” (Pet. Br.12) by interrogating detectives borne out by the record. Holland cites no reference to the death penalty whatever during his first three interrogations by the police. Rather, he relies exclusively on the interaction with LaRotunda, which lasted less than an hour before the ultimate videotaping of Holland’s statement, beginning on Monday morning at about 10:45 a.m. (H. Tr. 420-21, 469.) But despite extensive cross-examination, LaRotunda (whose testimony provides the only account of this conversation in the record) did not testify that he made repeated or persistent reference to the death penalty; indeed, he was never even asked whether he did so. Rather, he repeatedly testified only that either he or another officer who was present referred to the possibility that this could be a death penalty case (H. Tr. 600, 648), and that he “set up an analogy of like two coffee cups and a cigarette lighter and the analogy was that the cigarette lighter being the death penalty and the middle coffee cup being twenty-five years to life and the end coffee cup seven to ten years.” (H. Tr. 457.) The “analogy” is hardly “vivid” (Pet.Br.12)' — indeed, coffee cups and a cigarette lighter are rather homely items for conveying or symbolizing the terrifying reality of execution or imprisonment. While LaRotunda’s testimony is not especially clear about the role of the coffee and lighter in the “analogy,” the point seems to have been that it would be possible for Holland literally to take the more intimidating outcomes “off the table” by choosing to confess and seek a more lenient outcome. (H. Tr. 604.)

Surely, pointing out to a suspect that he might face execution if convicted, and that showing contrition by confessing might induce leniency, encourages the suspect to confess. And LaRotunda candidly acknowledged that that was precisely his goal in making the point. (H. Tr. 654-55.) But there is no evidence that LaRotunda misled Holland by falsely promising leniency, or put him under intense psychological pressure by unduly graphic descriptions of the pains of the death chamber. The trial court found that “LaRotunda made no promises or guarantees to the defendant at any time” (Decision and Order at 9), and Holland points to no evidence, indeed makes no claim, to the contrary. The state court did not unreasonably apply the Supreme Court’s test for voluntariness by finding, in the totality of the circumstances, that Holland’s statements were not rendered involuntary by LaRotunda’s (accurate) reference to the potential application of the death penalty or by his (accurate) implication that confession could lead to leniency. Indeed, the Second Circuit reached a similar conclusion in evaluating similar, arguably more aggressive or improper interrogation tactics. See Green v. Scully, 850 F.2d 894, 902-04 (2d Cir.1988).

Finally, Holland argues in the alternative that this is not merely a case of “unreasonable application of’ Supreme Court precedent, but one in which the state court’s decision was directly “contrary to” a Supreme Court decision, specifically, Reeves v. Alabama, 348 U.S. 891, 75 S.Ct. 214, 99 L.Ed. 700 (1954). Counsel’s diligent research is to be commended; while Reeves is more recent than Bram, it is far more obscure. The Supreme Court’s decision in this nearly half-century old case consists of just two words, “Judgment reversed,” and a pair of citations: “See Canty v. Alabama, 309 U.S. 629, 60 S.Ct. 612, 84 L.Ed. 988 [(1940)], and Vernon v. Alabama, 313 U.S. 547, 61 S.Ct. 1092, 85 L.Ed. 1513 [(1941)].” Canty and Vernon, in turn, are equally pithy summary reversals of findings of voluntariness, which rely primarily on the Court’s decision in Chambers v. Florida, 309 U.S. 227, 60 S.Ct. 472, 84 L.Ed. 716 (1940). Moreover, it appears that the Supreme Court has had no occasion to cite Reeves for any reason in the forty-eight years since it was decided.

In the absence of an opinion, it is impossible to tell what factors led the Court to find Reeves’ confession involuntary. Holland points to the fact, found in the opinion of the Supreme Court of Alabama, that Reeves had been told that “if he would admit having raped the prosecutrix that would keep him out of the electric chair.” Reeves v. State, 260 Ala. 66, 74, 68 So.2d 14 (1953). 4 But that opinion, in which the Alabama court affirmed Reeves’ conviction, gives a relatively innocuous account of the circumstances of his interrogation, and provides no insight into the Supreme Court’s reasoning. There is no reason to think that this promise alone produced the Supreme Court’s conclusion — and even if it did, Holland was made no such promise. Canty and Vernon, the two cases cited by the Supreme Court in Reeves, are similarly inscrutable, and in neither instance does the opinion of the state court shed much light on the Supreme Court’s holding. 5 Finally, Chambers, which establishes the principles applied summarily in all three cases, is entirely distinguishable from this case, as the following summary of its facts makes clear:

For five days petitioners were subjected to interrogations culminating in Saturday’s (May 20th) all night examination. Over a period of five days they steadily refused to confess and disclaimed any guilt. The very circumstances surrounding their confinement and their questioning without any formal charges having been brought, were such as to fill petitioners with terror and frightful misgivings. Some were practical strangers in the community; three were arrested in a one-room farm tenant house which was their home; the haunting fear of mob violence was around them in an atmosphere charged with excitement and public indignation. From virtually the moment of their arrest until their eventual confessions, they never knew just when any one would be called back to the fourth floor room, and there, surrounded by his accusers and others, in terrogated by men who held their very lives — so far as these ignorant petitioners could know — in the balance.- The rejection of petitioner Woodward’s first “confession”, given in the early hours of Sunday morning, because it was found wanting, demonstrates the relentless tenacity which “broke” petitioners’ will and rendered them helpless to resist their accusers further. To permit human lives to be forfeited upon confessions thus obtained would make of the constitutional requirement of due process of law a meaningless symbol.

Id. at 239-40, 60 S.Ct. 472 (footnote omitted). If the Supreme Court saw Reeves as a simple application of the Chambers precedent, it follows that this case is not controlled by Reeves.

Accordingly, the conclusion of the state court, applying the correct legal standard after a thorough hearing, that Holland’s confession was voluntary under the totality of the circumstances, was neither “contrary to” nor an “unreasonable application of’ Supreme Court precedent.

II. Miranda

As is apparent from the above discussion, even where the interrogation tactics utilized by the police do not rise (or sink) to the level of coercion sufficient to overbear the will of a normal criminal defendant or suspect, custodial interrogation of the sort to which Holland was subjected involves subtle and not-so-subtle psychological pressures that are at least to some degree coercive. In order to dispel that inherently coercive pressure, the Supreme Court in Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966), required police officers to follow a particular protocol before engaging in custodial interrogation — the famous Miranda warnings — to secure the admissibility of statements made in response to such interrogation. Compliance with Miranda and its progeny is required; such compliance also counters the coercive effect of routine interrogation practices such as those described above, making it far more difficult for a duly-warned defendant to claim that a confession was involuntary (thus incidentally reducing the need for the kind of case-by-case summary assessment of vol-untariness by the Supreme Court illustrated by decisions such as Reeves). In addition to his claim of involuntariness, Holland claims that his confessions should have been suppressed because the officers here did not comply with Miranda. This claim is also meritless.

The state court found that Holland was in custody from at least 9:15 p.m. on Sunday, September 1, when the investigating officers learned from Holland’s girlfriend that he had not returned home on the night of the murder. (Decision and Order at 15.) Holland was properly given Miranda warnings at that time. (Id. at 5.) In addition to requiring that warnings be given, however, Miranda also requires that the police “scrupulously honor[ ],” 384 U.S. at 479, 86 S.Ct. 1602, any invocation of the right to remain silent: “If the individual indicates in any manner, at any time prior to or during questioning, that he wishes to remain silent, the interrogation must cease. At this point he has shown that he intends to exercise his Fifth Amendment privilege; any statement taken after the person invokes his privilege cannot be other than the product of compulsion, subtle or otherwise.” Id. at 473-74, 86 S.Ct. 1602. See also Michigan v. Mosley, 423 U.S. 96, 101-06, 96 S.Ct. 321, 46 L.Ed.2d 313 (1975) (when suspect indicates he wishes to remain silent, interrogation must cease, but may under certain circumstances resume later). Holland argues that he invoked his right to remain silent by failing to answer and hitting his head on the table when confronted with his girlfriend’s failure to confirm his alibi, but that the police, after initially ceasing their questioning, returned to the attack early the next morning and renewed questioning.

The state court rejected this argument, holding that “[s]elective silences ... do not require suppression of later statements where, as here, the defendant understood his Miranda rights and nevertheless chooses later to speak with police.” (Decision and Order at 14.) As with the volun-tariness claim, Holland can only prevail if he can show that this ruling was “contrary to, or involved an unreasonable application of’ Supreme Court precedent. 28 U.S.C. § 2254(d)(1).

Holland cannot meet this standard. The basic principle he relies on, which respondents do not contest, is set forth in the passage from Miranda quoted above: once a suspect “indicates ... that he wishes to remain silent,” or “has shown that he intends to exercise his Fifth Amendment privilege,” interrogation must cease. 384 U.S. at 473-74, 86 S.Ct. 1602. The state trial court, however, ruled that Holland’s failure to respond when questioned at 9:15 p.m. about his girlfriend’s statement — his “selective silence[ ]” on that subject — after having been properly advised of his rights and choosing to answer earlier questions, and before choosing again to respond to renewed questioning the following morning, did not constitute an invocation of his privilege, or express a desire to remain silent.

Neither Miranda nor any subsequent Supreme Court decision directly confronts the issue of what kind of expression indicates a desire “to cut off questioning.” 384 U.S. at 474, 86 S.Ct. 1602. The Court’s decision in Mosley, however, demonstrates that the quoted passage from Miranda leaves room for reasonable differences of interpretation and application. Mosley unambiguously invoked his right to silence in response to police interrogation; the next morning, however, different officers came to question him about a different crime, again advised him of his rights, and secured his agreement to answer questions. The Supreme Court rejected the argument that Miranda’s command that “interrogation must cease” precluded the later questioning. The Court ruled that complying with the command to “cease” required more than “a momentary cessation of questioning,” but did not require “permanent immunity from further interrogation.” 423 U.S. at 102, 96 S.Ct. 321.

We can assume without deciding that, if Holland had asserted his right to cut off questioning, Mosley would not have permitted renewed questioning in this case, where the police did not seek to question Holland about an unrelated crime, and did not remind him of his right to silence, but simply returned to the attack on the very same subject on which he (on this hypothesis) had “exercise[d] his Fifth Amendment privilege.” Miranda, 384 U.S. at 473, 86 S.Ct. 1602. 6 But Mosley provides no sup port for Holland’s claim that he should be found to have asserted the privilege. Mosley, unlike Holland, had “declined to answer,” 423 U.S. at 105, 96 S.Ct. 321, “said he did not want to answer any questions about the robberies” about which the police questioned him, id. at 97, 96 S.Ct. 321, and “stated that he did not want to discuss the robberies,” id. at 104, 96 S.Ct. 321. Holland, in contrast, expressly waived his Miranda rights, answered numerous questions for an extended period, and, so far as the record reflects, never declined to answer, said he would not want to answer additional questions, or asked that the questioning stop. Instead, he simply failed to answer a particular question, and beat his head on the desk, presumably overcome by his emotions.

Of course, for Holland to invoke his right to silence, it was not necessary for him to incant any magic formula, like a witness before a congressional committee reading a statement prepared by counsel. It is enough, as Miranda teaches, for the suspect to “indicate! ] in any manner, at any time prior to or during questioning that he wishes to remain silent.” 384 U.S. at 473, 86 S.Ct. 1602. Does silence in the face of a particular question or questions constitute such an indication of “wishing] to remain silent,” or to “exercise his Fifth Amendment privilege”? Id.

The Second Circuit has clearly ruled that it does not. In United States v. Ramirez, 79 F.3d 298 (2d Cir.1996), a case strikingly similar to this one, the defendant had acknowledged receiving and understanding his Miranda warnings, and agreed to speak with officers. He “responded to two [questions] and was silent as to the third”; later, he answereu most” of the officer’s additional questions, but “did not respond to two” questions. Id. at 305. The Court rejected defendant’s claim that silence in response to these questions constituted an implied invocation of this right to silence, holding that his “silence in the wake of two questions, while answering others, did not constitute even an equivocal invocation of his right to remain silent.... Since Ramirez had been properly advised of his rights and agreed to answer questions, and thereafter neither remained entirely silent nor stated that he wished to be asked no further questions but instead simply answered some questions and did not respond to others, we agree with the district court’s conclusion that Ramirez did not invoke his right to remain silent, and that the agents’ further questions did not violate his Fifth Amendment rights.” Id. 7

If Ramirez represents a correct understanding of Supreme Court precedent— and this Court is of course compelled to recognize it is such — then the state court’s similar conclusion in the instant case certainly cannot be held to be an “unreasonable application” of that precedent. Accordingly, Holland’s second ground for seeking the writ must be denied.

III. The Sentence

Holland next argues that his sentence to life imprisonment without parole violates the due process and equal protection guarantees of the Fourteenth Amendment. While the claims are ultimately without merit, they are worthy of careful consideration, as they arise from a perhaps-unintended glitch in the New York capital murder statute.

The jury convicted Holland of first-degree murder, for which New York law authorizes capital punishment. 8 A sentence of death is of course not mandatory for that crime. See Woodson v. North Carolina, 428 U.S. 280, 288-301, 96 S.Ct. 2978, 49 L.Ed.2d 944 (1976). Instead, New York permits first-degree murderers to be sentenced to death, life imprisonment without parole, or a parolable indeterminate sentence of at least 20 years to life. 9 See N.Y. Penal L. §§ 60.06 (setting alternative dispositions for first-degree murder); 70.00(2)(a) (setting maximum term of indeterminate sentence); 70.00(3)(a)(i) (setting minimum term of indeterminate sentence). If the prosecutor seeks the death penalty, elaborate procedures are provided for determining whether capital punishment will be imposed. N.Y.Crim. Proc. L. § 400.27. The details of the procedure for assessing the aggravating and mitigating circumstances to be considered in the solemn life-or-death decision are not particularly relevant to this case. Two points, however, are critical to understanding Holland’s claim.

First, if the death penalty is sought, and the jury finds defendant guilty of first-degree murder, a separate sentencing proceeding is conducted, ordinarily before the same jury that found defendant guilty. N.Y.Crim. Proc. L. § 400.27(2). In such a case, the jury plays a dominant role in determining the sentence. If the jury unanimously determines beyond a reasonable doubt that the aggravating circumstance(s) in the case substantially outweigh the mitigating factor(s), and unanimously determines that the death penalty is appropriate, the court is required to impose a sentence of death (although the court may entertain a motion to set aside the sentence on specified grounds). N.Y.Crim. Proc. L. § 400.27(ll)(a), (d). If the jury unanimously determines that the sentence should be life imprisonment without parole, the court must impose that sentence. N.Y.Crim. Proc. L. § 400.27(ll)(e). If the jury is unable to reach a unanimous conclusion, sentencing reverts to the judge, but in that event, the judge is directed to sentence according to the ordinary indeterminate sentencing provisions (with a minimum term of 20 years to life), in effect excluding life without parole as an option. N.Y.Crim. Proc. L. § 400.27(ll)(c). In short, if the prosecutor seeks the death penalty, the jury alone can impose a sentence of death or life without parole; the highest sentence available if sentencing reverts to the judge is sentence of life imprisonment, with parole eligibility after 25 years.

Second, the prosecutor is given discretion, unconstrained by any provision of the capital sentencing statutes, to determine “at any time ... that the death penalty shall not be sought in a particular case.” N.Y.Crim. Proc. L. § 400.27(1). In that event, no special sentencing proceeding is required. Instead, upon a verdict or plea of guilty of first-degree murder, sentencing authority is vested in the judge, who may sentence the defendant either to life without parole or to an indeterminate term of no less than 20 years to life. In Holland’s case, the prosecutor elected not to seek execution. Accordingly this latter procedure applied, and the trial judge sentenced Holland to life without parole.

Holland raises two objections to this scheme. First, he contends that his sentence violates the principle announced by the Supreme Court in Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000), that “the Due Process Clause of the Fourteenth Amendment requires that a factual determination authorizing an increase in the maximum prison sentence for an offense ... be made by a jury on the basis of proof beyond a reasonable doubt.” 530 U.S. at 469, 120 S.Ct. 2348 (quoted in Pet. Br. 32). Second, he argues that the sentencing scheme violates equal protection, by arbitrarily permitting defendants for whom the death penalty had been sought to be sentenced to life without parole only by a unanimous jury (prohibiting a judge from imposing that punishment on his own authority), while withholding a jury determination of the appropriateness of life without parole from defendants for whom the death penalty is not sought (permitting the judge to impose that sentence under those circumstances).

The first issue is easily disposed of. Apprendi has no application to this case. In Apprendi the Supreme Court held that it is unconstitutional for a state to permit a judge to impose a sentence above what would otherwise be the statutory maximum for the crime found beyond a reasonable doubt by a jury, on the strength of aggravating factors found by the judge alone on a lesser standard of proof. Accordingly, the Court reversed a twelve-year sentence for possession of a firearm with an unlawful purpose. While that offense, to which Apprendi pled guilty, ordinarily carries a maximum sentence of ten years’ imprisonment, the sentence was enhanced, under the state’s hate crime law, based on the judge’s factual finding by a preponderance of the evidence that the crime had been intended to intimidate the victim on account of race. Id. at 468-70, 120 S.Ct. 2348. The Court held that since this factual finding “increase[d] the penalty for [the] crime beyond the prescribed statutory maximum,” due process required the finding to be made by a jury and proved beyond a reasonable doubt. Id. at 490,120 S.Ct. 2348.

Holland’s case does not implicate the constitutional concerns addressed in Ap-prendi. A jury found Holland guilty of first-degree murder, having properly found all the elements of that offense beyond a reasonable doubt, under instructions not challenged here. Life without parole, the sentence imposed on Holland, is expressly authorized as a punishment for the crime found by the jury, N.Y. Penal L. § 60.06, without the need for any additional fact-finding by the judge. 10 There is therefore no basis for a claim that Holland’s rights to a jury trial or to conviction by proof beyond a reasonable doubt, as interpreted in Apprendi, were violated here. 11

Holland’s second point is less easily brushed aside. Life imprisonment without any hope of parole or other release is a particularly harsh sentence, thought by some to be a fate as bad as, or possibly even worse than, death itself. See, e.g., Holman v. Page, 95 F.3d 481, 487 (7th Cir.1996) (“Natural life imprisonment is a stern punishment, for some perhaps worse than death.”); People v. Bloom, 48 Cal.3d 1194, 1223 n. 7, 259 Cal.Rptr. 669, 774 P.2d 698 (1989) (“While qualitatively different from the death penalty, the punishment of life imprisonment without hope of release has been regarded by many as equally severe.” (citing philosophers)). 12 As Holland characterizes the statute, New York reserves this extreme punishment for a very narrow class of offenders. No crime other than first-degree murder — a limited category of selected, highly aggravated intentional killing — exposes an offender to this punishment. Moreover, in cases where capital punishment is sought by the prosecution, life without parole is treated as solemnly, and with as many procedural safeguards, as the death penalty. In such cases, only a unanimous jury may impose that punishment; if the jury fails to agree that life without parole is called for, the sentencing authority reverts to the judge, but the judge is forbidden to preclude parole, and may impose at most a life sentence that is subject to parole after 25 years.

In contrast, however, if the prosecutor does not elect to seek the death penalty, the defendant loses the benefit of having a jury assess the need for imprisonment without hope of release, and such a sentence may be imposed unilaterally by a judge. As Holland points out, there is no clear criterion for determining which offenders fall into which category; the prosecutor’s unguided discretion determines who is subject to which procedure. Moreover, to the extent that one assumes there is a systematic difference between the two groups, that difference presumably is that the offenders for whom the death penalty is sought are determined by the prosecutor to be more culpable and/or more dangerous; that’s why the prosecutor seeks to have them executed. It would appear odd to make it harder to subject these defendants to unparolable life sentences than others, whom the prosecutor has not deemed worthy of the ultimate penalty.

Accordingly, Holland argues, to the extent that New York permitted him to be sentenced so severely by a judge, while granting the additional protection of jury sentencing before the same sentence could be imposed on other offenders convicted of the same offense, who are not rationally distinguishable from him (except to the extent that they are arguably even more deserving of the penalty), it denied him the equal protection of the law in violation of the Fourteenth Amendment.

Looked at in this way, it is indeed difficult to see what rational purpose is served by New York’s scheme. If death is sought by the prosecutor, the maximum punishment available to the judge, absent a unanimous jury vote for life without parole, is life with the possibility of parole after 25 years; if the prosecutor does not seek death, the judge can impose life without parole on his own authority. One is hard put to understand why this distinction would be created; tellingly, respondents make no attempt even to hypothesize a rational policy reason for the distinction. See also William C. Donnino, Practice Commentary to N.Y. Penal L. § 125.27, at 395 (McKinney 1998) (arguing that “[t]he logic of the distinction is questionable”).

Holland does not argue that the Constitution requires that sentences of life without parole be imposed only by a jury. Indeed, even capital sentencing by a jury, though it has a venerable history in the United States, is not constitutionally required. See, e.g., Harris v. Alabama, 513 U.S. 504, 509-10, 115 S.Ct. 1031, 130 L.Ed.2d 1004 (1995); Spaziano v. Florida, 468 U.S. 447, 465, 104 S.Ct. 3154, 82 L.Ed.2d 340 (1984). If the legislature sees life without parole as tantamount to a death sentence, the legislature might choose to subject that penalty to the same procedural safeguards as capital punishment, but if it sees the sentence as simply the highest on a range of comparable prison terms, not radically different from parolable life terms and divided by a major moral chasm from execution, it could lawfully choose to allow the judge to impose life without parole after normal sentencing procedures. Holland argues, however, that the legislature must choose one option or the other for all first-degree murderers who have been spared the death penalty, and that there is no rational basis for applying different procedures for imposing life without parole according to whether the defendant was spared death by a jury that fails to reach a unanimous verdict or by prosecutorial discretion.

Nevertheless, the peculiarity of this wrinkle in New York’s statute does not entitle Holland to grant of the writ. To begin with, the Court must be mindful of the standard of review. As repeatedly noted, where the state courts have rejected a constitutional claim, a habeas court may not grant relief simply because it believes constitutional error has been committed. Habeas relief is only available when the state court’s decision is contrary to, or an unreasonable application of Supreme Court precedent. 28 U.S.C. § 2254(d)(1). This standard applies when the state court (1) disposes of a federal claim on the merits and (2) reduces its disposition to judgment — even if it does not explicitly refer to the federal claim or to relevant federal case law. Sellan v. Kuhlman, 261 F.3d 303, 314 (2d Cir.2001). While Holland argues to the contrary (Pet. Br.32), the state courts clearly denied this claim on the merits: the argument was clearly presented in Holland’s brief to the Appellate Division, Resp. Ex. A at 36-42; the prosecution did not argue that the claim had been procedurally defaulted; and the Appellate Division ruled that those of Holland’s arguments not specifically addressed in its opinion were “without merit.” People v. Holland, 268 A.D.2d 536, 703 N.Y.S.2d 57, 58 (2d Dep’t 2000).

The rejection of Holland’s constitutional argument is not directly contrary to any Supreme Court decision; neither party cites, nor has the Court found, any Supreme Court case that addresses the equal protection implications of distinctions between judge and jury sentencing. The question thus becomes whether the state court’s rejection of Holland’s argument was an “unreasonable application” of Supreme Court precedent. In this case, the parties agree that the principle to be applied is the extremely lenient rational basis test applied by the Supreme Court in assessing claims of denial of equal protection not involving “discrete and insular minorities.” United States v. Carolene Products Co., 304 U.S. 144, 152 n. 4, 58 S.Ct. 778, 82 L.Ed. 1234 (1938). 13

Difficult as it may be to explain in isolation the narrow classification Holland challenges, the distinction drawn by the statute is certainly not irrational when looked at in its full context within the New York capital sentencing system. The procedures established for death and non-death cases in New York are quite different, and the sentence of life without parole plays a somewhat different role in the distinct situations of a capital sentencing trial, with its special procedures, and a non-capital sentencing.

First, where capital punishment is sought, the jury plays a unique role in sentencing. That role is longstanding in the tradition of American law. Since Tennessee first abolished mandatory capital sentencing for murder in 1838, and vested discretion in juries to impose a non-capital sentence, the jury has been the dominant institution in capital sentencing. In that context, it is hardly surprising that the jury is permitted to impose life without parole, in appropriate eases, as one of its sentencing options. The sentence of life without parole is not specially reserved to juries in such cases; rather, the capital sentencing decision is reserved to the jury, and the possibility of imposing a sentence of life imprisonment without parole is assigned to it as an adjunct alternative as part of the jury’s special responsibility in administering capital punishment.

New York’s special jury sentencing procedure for capital offenses is not particularly unusual; indeed, it is a version of the predominant practice in states that provide for capital punishment. The procedure ultimately derives from the capital sentencing procedure suggested by the American Law Institute’s Model Penal Code in 1962. Model Penal Code & Commentaries, pt. II, § 210.6 (Official Draft and Revised Comments 1980). At that time, most states that utilized the death penalty left the capital sentencing decision to juries, who had total discretion to impose or withhold the penalty as they saw fit. See id. pt. II, § 210.6, cmts. 4(c), 7, at 131,142. The Model Penal Code’s proposal sought to guide the discretion of juries by providing for the introduction of evidence at a bifurcated sentencing procedure, and for instructions directing juries to consider specified aggravating and mitigating circumstances in deciding the appropriateness of capital punishment. See id. pt. II, § 210.6, cmts. 5-6, 8. Although this model was not immediately influential, it came to be adopted and adapted by many states in the wake of the Supreme Court’s invalidation of standardless death penalty sentencing as cruel and unusual in Furman v. Georgia, 408 U.S. 238, 92 S.Ct. 2726, 33 L.Ed.2d 346 (1972), and its basic approach was approved by the Court as constitutionally appropriate in Gregg v. Georgia, 428 U.S. 153, 96 S.Ct. 2909, 49 L.Ed.2d 859 (1976). The special jury sentencing procedure is thus a feature specifically tied to capital cases.

Where death is not in question, however, sentences are usually imposed by judges, not juries, without the special elaborate procedures provided in capital cases. Even the extreme sentence of life without parole is available for a variety of crimes in American jurisdictions, without departure from the basic sentencing procedures provided for other non-capital sentences. 14 Accordingly, there is nothing suspicions, irrational, or even unusual about permitting a judge to impose a sentence of imprisonment — even life imprisonment without the possibility of parole — where that sentence is provided for by law and proportionate to the offense, while providing a special set of procedures for capital cases.

Second, the sentence of life without parole plays a different role in the distinct situations of a capital sentencing trial and a non-capital sentencing for first-degree murder. In the capital context, providing the jury with the alternative of life without parole mitigates the likelihood that the death penalty will be imposed. Studies have shown that juries are less likely to impose capital punishment when they know that they can fully protect the public by imposing life without parole than when they fear that failure to impose the death sentence risks future release of a dangerous murderer. See, e.g., Theodore Eisen-berg & Martin T. Wells, Deadly Confusion: Juror Instructions in Capital Cases, 79 Cornell L.Rev. 1, 7 (1993) (“Not surprisingly, jurors assessing dangerousness attach great weight to the defendant’s expected sentence if a death sentence is not imposed.... [JJurors who believe the alternative to death is a relatively short time in prison tend to sentence to death. Jurors who believe the alternative treatment is longer tend to sentence to life.”). Giving the jury the option of imposing that sentence, with the same stringent requirements of unanimous decision-making and an affirmative vote to require life without parole that apply to the death sentence, serves the benevolent purpose of restricting the death penalty to cases where the jury believes that the ultimate sanction is absolutely necessary. Since the New York death penalty statute is particularly narrow and restrictive, and was finally adopted in 1995 after more than twenty years of political controversy and de facto abolition, the legislature presumably intended this result.

In cases in which the death penalty is not sought, however, permitting the judge to impose life without parole, in his or her discretion, may serve a different set of purposes, which also relate to mitigating the capital punishment regime. The statute emphasizes that prosecutors always have broad discretion not to seek the death penalty, and explicitly links the power of the judge to sentence to life without parole to that discretion. See N.Y.Crim. Proc. L. § 400.27(1) (“Nothing in this section shall be seemed to preclude the people at any time from determining that the death penalty shall not be sought in a particular case, in which case ... the court may sentence such defendant to life imprisonment without parole [or to a parola-ble life term]”). The Governor’s explanation of this aspect of the proposed statute stated: “To safeguard the public, the bill provides that when prosecutors do not seek the death penalty, defendants can still be sentenced to life without parole.” Governor’s Memorandum of Approval of L.1995, c. 1, Mar. 7, 1995, reprinted in Historical and Statutory Notes to N.Y. Correct. L. art. 22-B, at 244 (McKinney Supp.2001-02) (“Governor’s Signing Statement”). The availability of life without parole in this circumstance evidently represents a trade-off for the liberal availability of prosecutorial discretion, provided as a hedge against the concern that prosecu tors exercising this discretion might fail to be sufficiently zealous in protecting the public; in exchange for giving the prosecutor an unbridled option to exercise mercy, the state requires that the judge have an option of guaranteeing public safety by imposing an unparolable life term. 15

In light of the difficult choices presented to a legislature concerned to make capital sentences available, but limit them to only the fewest, most compelling eases, the procedures adopted make much more sense. Indeed, the only peculiar aspect of the entire scheme is that life without parole is not provided as an option to the judge in cases where the jury is unable to agree on either death or life without parole. But even this is not inexplicable. Life without parole is still a sentencing option in death penalty cases; as noted above, it is an option presented to the jury, in large part to remind them that death is not the only option for guaranteeing public safety. If New York’s system functions as intended, prosecutors will seek the death sentence only in the most extreme cases, in which a plausible case can be made for the ultimate sanction, and in which it can be expected that juries will frequently find life without parole a desirable option. Where the jury has rejected this penalty, or failed to agree on it, the prosecutor has had the chance, on behalf of the public, to seek that sentence. The legislature could reasonably find that there is no need to allow the prosecutor a chance to repeat his unsuccessful argument to another decisionmaker. Where the prosecutor has not sought a capital sentence, on the other hand, life without parole has not been rejected either by the prosecutor or by a jury, and under these circumstances it is a reasonable legislative choice to give the prosecution an opportunity to seek the imposition of this penalty.

Looked at in isolation, as Holland would have us do, the inconsistency in procedure to which he calls attention is not susceptible to elegant rationalization. But legislation is a product of compromise, and courts are appropriately reluctant to find violations of fundamental constitutional principle in the incongruities that sometimes result. Cf. Metropolitan Life Ins. Co. v. Ward, 470 U.S. 869, 898, 105 S.Ct. 1676, 84 L.Ed.2d 751 (1985) (“Rational basis scrutiny does not require that the classification be mathematically precise ... [or] ‘demand a surveyor’s precision’ in fashioning classifications”). Taken as a whole, and in context, the New York legislature was not irrational in providing a life without parole sentencing option both where the jury determines the sentence (because the death penalty is in play) and where sentencing is for the judge (because it is not). Each sentencing procedure necessarily operates differently, but the options provided within each sentencing system are rational in light of the different goals and problems pursued in death and non-death cases. If, as a consequence of the different legislative choices made for each procedural setting, inconsistencies develop between what would appear to be analogous functions in the different contexts, even if it is possible to imagine alternative solutions that might resolve the inconsistency without apparent damage to the policies being pursued, that does not mean that the legislature’s choices are not rationally related to the important and difficult public policies it is pursuing.

At a minimum, given the constraints on habeas corpus relief, this Court cannot say that the state courts’ decision to uphold the New York procedure against equal protection challenge was an unreasonable application of the Supreme Court’s established equal protection principles. Accordingly, relief will not be granted on this ground.

IV. Jury Selection

Finally, Holland argues that the trial court erred in handling his objection, pursuant to Batson v. Kentucky, 476 U.S. 79, 106 S.Ct. 1712, 90 L.Ed.2d 69 (1986), that the prosecutor engaged in racial discrimination in exercising a peremptory challenge against a black potential juror. This claim was presented to the state court on appeal and rejected on independent and adequate state procedural grounds. Accordingly, it has been procedurally defaulted and may not be the basis for habeas relief.

On appeal to the Appellate Division, Holland argued that the prosecution had exercised its peremptory challenges in a discriminatory manner by excluding four black venirepersons, and that the prosecutor had either failed to present reasons (as to one juror) or had given inadequate and pretextual reasons (as to the other three). (Resp. Ex. A at 25-29.) The Appellate Division rejected the claim, noting that “the prosecutor stated his reasons for the record and the defendant did not indicate dissatisfaction with those explanations. Thus, the precise Batson issue raised on appeal, i.e., that the proffered reasons were pretextual, was not preserved for appellate review.” 703 N.Y.S.2d at 58; Resp. Ex. C at 2. In support of its conclusion, the court cited People v. Childress, 81 N.Y.2d 263, 268, 598 N.Y.S.2d 146, 614 N.E.2d 709 (1993), in which the New York Court of Appeals required that trial counsel, “in order to give the trial court a proper foundation to evaluate [a Batson ] claim — as well as to ensure an adequate record for appellate review — ... should articulate and develop all of the grounds supporting the claim, both factual and legal, during the colloquy in which the objection is raised and discussed.”

A federal court may not review a state court decision that rests on an adequate and independent state procedural default unless petitioner can show both cause and prejudice or a fundamental miscarriage of justice. See Coleman v. Thompson, 501 U.S. 722, 749-50, 111 S.Ct. 2546, 115 L.Ed.2d 640 (1991); Fama v. Commissioner of Correctional Services, 235 F.3d 804, 809 (2d Cir.2000). In order for federal review to be barred, “ ‘[t]he state court must actually have relied on the procedural bar as an independent basis for its disposition of the case.’ ” Fama, 235 F.3d at 809 (quoting Harris v. Reed, 489 U.S. 255, 261-62, 109 S.Ct. 1038, 103 L.Ed.2d 308 (1989)). “Because it can be ‘difficult to determine if the state law discussion is truly an independent basis for decision or merely a passing reference,’ such reliance on state law must be ‘clear from the face of the opinion.’ ” Id. (cita tions omitted). In the recent case of Galarza v. Keane, 252 F.3d 630, 637 (2d Cir.2001), the Second Circuit rejected a claim of procedural bar with respect to a similar Batson challenge, because the Appellate Division “explicitly and only addressed the merits,” despite an arguable procedural default. Here, however, the opposite is the ease: the Appellate Division “explicitly and only” addressed the procedural issue. Without engaging in any discussion of the merits, the court simply held that the issue being raised by Holland “was not preserved for appellate review.” 703 N.Y.S.2d at 58; Resp. Ex. C at 2.

In this Court, Holland has changed tactics slightly, focusing his discussion entirely on the one potential juror, one Mary Williams, as to whom the prosecutor gave no reasons at all for his peremptory challenge. Apparently, Holland contends that the Appellate Division’s procedural default ruling, which referenced the defendant’s failure to object to the prosecutor’s explanations for his challenges, did not apply in this instance. (Pet.Br.42.) But this maneuver can’t avoid the procedural default. First, Holland never challenged the correctness of the Appellate Division’s finding of default under state procedural law, either by rehearing petition in that court or in his submissions seeking leave to appeal to the Court of Appeals. Second, the factual context of this particular challenge serves to highlight the rationale for the Appellate Division’s procedural ruling.

Williams was one of the first two jurors peremptorily challenged by the prosecution, both of whom were black. Defense counsel immediately objected, focusing his complaint not on Williams, but on the other challenged juror:

MR. TRAYNOR: Your Honor, I don’t like the Assistant District Attorney getting rid of Mr. Pettigrew. He is a black man and there is nothing wrong with the guy. I don’t know why the Assistant District Attorney is exercising his peremptory challenge like that. I’m suspicious about it in terms of the racial composition of this trial, everybody being white.

Resp. Ex. J (“Voir Dire Tr.”) at 229-30. After a brief discussion of the race of the already-seated juror, 16 the judge, without explicitly deciding whether the defense had “made a prima facie showing that the circumstances give rise to an inference that a member of the venire was struck because of his race,” Galarza, 252 F.3d at 636; see Batson, 476 U.S. at 96-98, 106 S.Ct. 1712, proceeded immediately to require the prosecutor to “set forth some known [sic; presumably a transcriber’s error for ‘non-’] racial reasons” for the challenge. At this point, defense counsel interjected: “I’m raising a Batson challenge. They are two black people. I think there is something wrong in terms ... of motivation by the prosecutor being racial.” Voir Dire Tr. 230 (emphasis added). There ensued an extensive discussion as to whether the challenges betrayed a pattern that suggested discrimination, and of the reasons for the challenge to Mr. Pettigrew, which the judge ultimately found persuasive. Voir Dire Tr. 232-33. 17 After the judge ruled that the prosecutor had set forth adequate non-discriminatory reasons for challenging Pettigrew, defense counsel said, “The other lady also, the last person also. I reiterate, your Honor, I suspect motivation based on racial membership.” Voir Dire Tr. 233. The judge simply replied, “That is denied.” Voir Dire Tr. 233.

Undoubtedly, the trial judge could have been more careful. He did not, as recommended in Batson, initially consider whether the defense had made a prima facie case of discrimination, although he briefly expressed doubt on that score: “Well, if juror number one is not Caucasian, then I think you would have a very difficult time convincing me there is a pattern.” Voir Dire Tr. 231. Nevertheless, he proceeded to request a reason for the peremptory challenge of the juror on whom defense counsel had primarily focused. The judge then appeared to brush aside defense counsel’s less emphatic reminder, at the end of the discussion, that he also questioned the challenge of Ms. Williams. But the defense did not make an adequate record for the Appellate Division to consider .the Batson issue. Although the prosecution eventually used 14 peremptory challenges, eight of them against African-Americans (three of the twelve jurors ultimately seated were black), and the defense sought and received explanations for a number of those challenges, the defense never requested (and the court never made) a finding that a prima facie case of discrimination had been made, and the defense never returned to the issue of the challenge to juror Williams.

The Appellate Division’s ruling of procedural default was clearly directed to the entire Batson issue presented by Holland on direct appeal, and rightly so. Far from developing “all of the grounds supporting the claim,” as required by the Childress decision, defense counsel was content simply to tack on a mild objection to the Williams challenge to the far more vigorous contest over Pettigrew. On appeal, the four allegedly discriminatory challenges were attacked jointly, and when the Appellate Division ruled the issue unpre-served, the defense did not question that ruling. The issue having been defaulted, Holland cannot now revive it by suddenly highlighting the Williams challenge as a distinct legal issue.

Finally, it should be noted that even a procedurally defaulted issue can be raised if the “petitioner can show both cause and prejudice or a fundamental miscarriage of justice.” Fama, 235 F.3d at 809. 18 Neither in his original brief nor in his reply brief (filed after respondents had clearly and emphatically raised the procedural default issue) does Holland even attempt to show either, and no apparent basis for arguing either appears in the record.

Accordingly, Holland’s Batson claim is denied as procedurally defaulted.

CONCLUSION

Accordingly, for the reasons set forth above, the petition for habeas corpus is denied. As petitioner has not made a substantial showing of the denial of a constitutional right with respect to the first, second, and fourth claims, or with respect to the Apprendi challenge to his sentencing, a certificate of appealability will not issue as to those issues. See 28 U.S.C. § 2253(c)(2); Lucidore v. New York State Div. of Parole, 209 F.3d 107, 111-13 (2d Cir.2000).

In the Court’s view, however, the equal protection challenge to the different procedures for imposing life without parole in cases in which the prosecutor does or does not seek the death penalty presents a novel issue and provides fair ground for litigation regarding a constitutional challenge to a significant aspect of an important New York statute. The issue has not been addressed by the state or federal courts, and was not explicitly analyzed by the state courts in this case, even though the defendant properly briefed and preserved the issue. Given these circumstances, a certificate of appealability limited to this issue is granted.

SO ORDERED.

1 . Initially, the officers sought and obtained both Holland’s-and his girlfriend’s consent to search the apartment. But Holland, after orally consenting, refused to sign the consent form, and the girlfriend, who did sign a consent, gave a home address different from the apartment, creating some ambiguity about whether she continued to live there or had moved out due to difficulties in the relationship. Apparently due to questions about whether these consents would be sufficient, the officers commendably sought a warrant, at the cost of considerable additional effort and some additional delay.

2 . Beyond question, the Appellate Division rejected this claim on the merits. State court decisions have been found to be "on the merits" when the court said no more than "denied.” Sellan v. Kuhlman, 261 F.3d 303, 314 (2d Cir.2001); see Jenkins v. Artuz, 294 F.3d 284, 291-92 (2d Cir.2002) (disposition is "on the merits" when court states simply that claim is "without merit”). Here, the state court specifically ruled that the record supported the trial court's determination that the confession had not been the product of "coercive police strategy.” 703 N.Y.S.2d at 58; Resp. Ex. C at 2.

3 . See, e.g., Andresen v. Maryland, 427 U.S. 463, 472 & n. 6, 96 S.Ct. 2737, 49 L.Ed.2d 627 (1976) (noting that "continued validity of the broad statements contained in some of the Court's earlier cases,” including Boyd, "ha[s] been discredited by later opinions”). See also George C. Thomas III, When Constitutional Worlds Collide: Resurrecting the Framers’ Bill of Rights and Criminal Procedure, 100 Mich. L.Rev. 145 (2001) (persuasively recounting how the Warren Court's incorporation of Bill of Rights guarantees into the Fourteenth Amendment led to the loosening of restrictions on federal law enforcement officers while tightening those on state officers, and less persuasively arguing for strengthening limits on federal officers today).

4 . It didn’t. Reeves was sentenced to death.

5 . In affirming the defendant’s conviction in Canty, the Alabama Supreme Court summarily concluded that his written confession was voluntary and admissible, noting testimony that the defendant had been well-treated while in custody. Canty v. State, 238 Ala. 384, 191 So. 260, 262-63 (1939). In Vernon, the Alabama Supreme Court again concluded that the defendant’s confession was voluntary, noting that the defendant had failed to impeach the state’s evidence of voluntariness, "not even producing] a conflict in such testimony.” Vernon v. State, 239 Ala. 593, 196 So. 96, 100 (1940). As in Reeves, the state court’s opinion does not provide details of the circumstances in which the confession was given.

6 . The Mosley Court identified four factors that permitted the renewed questioning:

This is not a case ... where the police failed to honor a decision of a person in custody to cut off questioning, ... by persisting in repeated efforts to wear down his resistance and change his mind. In contrast to such practices, the police here [1] immediately ceased the interrogation, [2] resumed questioning only after the passage of a significant period of time and [3] the provision of a fresh set of warnings, and [4] restricted the second interrogation to a crime that had not been a subject of the earlier interrogation.

423 U.S. at 105-06, 96 S.Ct. 321. This case does not require us to address whether it would be a "reasonable application” of Mosley to conclude that renewed questioning would be permissible, after a clear invocation of the right to silence, where only two of these factors are present — where, as here, the police immediately ceased questioning for a significant period of time, but then returned to question on the same subjects without renewed warnings.

7 . Holland argues that the officers themselves must have understood his actions as indicating his unwillingness to be questioned, since they responded by stopping the interrogation. (Pet.Br.24-25.) First, as a factual matter, this inference is by no means compelling. The officers' response might well have been tactical, based on a judgment that allowing Holland (who was clearly troubled by hearing that his girlfriend had not backed up his alibi) to ponder his situation would be more likely to produce a confession than further questioning, which might have led to continued denials or an invocation of rights. Second, as a matter of law, the subjective belief of the officers as to the meaning or the legal implications of Holland’s actions is not controlling nor even relevant to the issue before the Court, which turns on whether a reasonable officer would have taken his actions as invoking his rights, and not on what the particular officers involved believed. See Mosley, 423 U.S. at 105, 96 S.Ct. 321 (concluding that officer’s subsequent interrogation about unrelated homicide "was quite consistent with a reasonable interpretation of Mosley’s earlier refusal to answer any questions about the robberies”); cf. United States v. Dhinsa, 171 F.3d 721, 725 (2d Cir.1998) (”[W]e judge the reasonableness of an officer's actions [under the Fourth Amendment] based on the objective circumstances surrounding her actions and not on her subjective intent.”).

8 . Section 125.27 of the New York Penal Law, substantially overhauled in 1995 in an effort to restore a constitutional death penalty, defines first-degree murder as intentionally causing the death of another person under circumstances including one or more of thirteen aggravating circumstances, including that the victim was killed while the defendant was engaged in robbery or certain other crimes. See N.Y. Penal L. § 125.27(l)(a)(vii).

9 . More precisely, the sentence will be an indeterminate term, with a required maximum of life, and a required minimum of no less than 20 years, and no more than 25 years.

10 . In contrast, imagine a statute under which, following a jury verdict convicting defendant of a crime defined simply as intentional killing and punishable by a sentence up to a parolable life term, a judge was nevertheless authorized to impose a sentence of life without parole based on the judge’s finding, by a preponderance of the evidence, of certain ag gravating circumstances. Such a statute would raise serious questions under Apprendi, since the defendant would be subjected based on judicial fact-finding to a sentence more severe than that authorized based on the facts found by the jury.

11 . Respondents question whether Holland's Apprendi claim was properly exhausted before the state courts. (Resp.Br.15-16.) There is no need to decide whether they are correct. Exhaustion is a prerequisite to the grant of the writ, 28 U.S.C. § 2254(b)(1), not to its denial. Under 28 U.S.C. § 2254(b)(2), "An application for a writ of habeas corpus may be denied on the merits, notwithstanding the failure of the applicant to exhaust the remedies available in the courts of the State.”

12 . But cf. People v. Memro, 11 Cal.4th 786, 879, 47 Cal.Rptr.2d 219, 905 P.2d 1305 (1995) (holding that prosecutor's comment to jury during sentencing that life imprisonment without parole was " 'legally not worse' than death was accurate as a legal matter, whatever philosophical feelings individuals might have on the subject' ” (citing Bloom, 48 Cal.3d at 1223 n. 7, 259 Cal.Rptr. 669, 774 P.2d 698)).

13 . Holland himself proposes that the New York scheme be judged according to whether it is "rationally related to a legitimate state purpose.” (Pet. Br. 35-36, citing Near v. Minnesota, 283 U.S. 697, 51 S.Ct. 625, 75 L.Ed. 1357 (1931); Mayer v. Chicago, 404 U.S. 189, 92 S.Ct. 410, 30 L.Ed.2d 372 (1971); and M.L.B. v. S.L.J., 519 U.S. 102, 122, 117 S.Ct. 555, 136 L.Ed.2d 473 (1996)). The Supreme Court has on occasion suggested that heightened scrutiny is required where legislative classifications affect a "fundamental interest.” E.g., Dunn v. Blumstein, 405 U.S. 330, 335-42, 92 S.Ct. 995, 31 L.Ed.2d 274 (1972); Shapiro v. Thompson, 394 U.S. 618, 638, 89 S.Ct. 1322, 22 L.Ed.2d 600 (1969). The interest in personal liberty, and more specifically of avoiding life imprisonment without parole, certainly seems fundamental. However, the Supreme Court also has suggested that the category of "fundamental interests” is limited to interests "explicitly or implicitly guaranteed by the Constitution.” San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 33, 93 S.Ct. 1278, 36 L.Ed.2d 16 (1973). Absent a so-called “suspect” classification or one affecting a fundamental interest, state legislative classifications within the criminal law are owed particular deference. See Pennsylvania v. Ashe, 302 U.S. 51, 55, 58 S.Ct. 59, 82 L.Ed. 43 (1937) ("Save as limited by constitutional provisions safeguarding individual rights, a State may choose means to protect itself and its people against criminal violation of its laws. The comparative gravity of criminal offenses and whether their consequences are more or less injurious are matters for its determination.”). "[Pjreventing and dealing with crime is much more the business of the States than it is of the Federal Government,” and accordingly, "we should not lightly construe the Constitution so as to intrude upon the administration of justice by individual States.” McMillan v. Pennsylvania, 477 U.S. 79, 85, 106 S.Ct. 2411, 91 L.Ed.2d 67 (quoting Patterson v. New York, 432 U.S. 197, 201-02, 97 S.Ct. 2319, 53 L.Ed.2d 281 (1977)). At any rate, the matter of levels of equal protection scrutiny is sufficiently vexed that a state court’s acceptance of the standard proposed by petitioner himself could not easily be called "unreasonable.”

14 . See, e.g., 18 U.S.C. § 3559(c) (mandating life imprisonment for "serious violent felonies'' if defendant previously has been convicted of at least two "serious violent felonies" or at least one "serious violent felony” and at least one "serious drug offense”); 21 U.S.C. § 848 (authorizing life imprisonment for engaging in continuing criminal enterprise, and mandating life imprisonment for principal administrators, organizers, and leaders of continuing criminal enterprises); U.S.S.G. § 2A1.1 (setting base offense level of life imprisonment for first-degree murder); 11 Del.Code Ann. § 4214 (authorizing life imprisonment without parole for certain repeat offenders convicted of specified felonies); Ga. Code Ann. § 17-10-7 (mandating life imprisonment without parole for certain repeat offenders convicted of certain "serious violent felonies”); Haw.Rev.Stat. § 706-656(1) (mandating life imprisonment without parole for individuals convicted of first-degree murder or first-degree attempted murder); Ind. Code Ann. § 35-50-2-8.5 (authorizing life imprisonment without parole for certain repeat offenders convicted of certain serious felonies); Iowa Code Ann. §§ 902.1 (mandating life imprisonment for individuals convicted of "Class A” felonies and explicitly removing such individuals from eligibility for parole unless sentence is commuted to term of years by governor); La.Rev.Stat. Ann. § 15:529.1 (authorizing life imprisonment without parole for certain repeat offenders convicted of certain serious felonies or "crimes of violence”); 17-A Me.Rev.Stat. Ann. §§ 1201(A), 1251 (authorizing life imprisonment for individuals convicted of murder and explicitly exempting such individuals from eligibility for split sentences consisting of imprisonment and probation); Mass. Gen. Laws Ann. ch. 265, § 2 (exempting individuals sentenced to life imprisonment for first-degree murder from eligibility for parole); Mich. Comp. Laws Ann. ¶ 791.234(4) (exempting individuals sentenced to life imprisonment for first-degree murder from eligibility for parole); N.C. Gen. Stat. § 14-7.7 (authorizing life imprisonment without parole for certain repeat offenders convicted of violent felonies); 42 Pa. Cons. Stat. § 9714 (authorizing life imprisonment without parole for certain repeat offenders convicted of certain specified "crimes of violence”); R.I. Gen. Laws § 1 l-23-2(b) (autho rizing life imprisonment without parole, at court's discretion, for individuals convicted of first-degree murder under certain conditions). But see W. Va.Code § 62-3-15 (providing for jury determination of whether defendant found guilty by jury of first-degree murder should be eligible for parole); State v. LaRock, 196 W.Va. 294, 312-15, 470 S.E.2d 613 (1996) (under W. Va.Code § 62-3-15, court may bifurcate jury trial into guilt phase and “mercy” phase in its discretion).

15 . Holland emphasizes that the legislative history cited above goes on to point out that in non-death-penalty cases, the state will also save money by obviating the special jury proceeding. He claims that this suggests that the true motivation for vesting the imposition of life without parole in the judge rather than the jury is simply to save money. Pet. Br. 36; see Governor's Signing Statement at 244 ("In these cases, moreover, the court sentences the defendant; the expenses attendant to sentencing proceedings are thus avoided.”). However, this mischaracterizes the state’s scheme. What needs explanation is not the "elimination” of jury sentencing — for sentencing by the judge is the norm where the death penalty is not an option — but rather the addition of life without parole to the judge’s arsenal in these cases. The expense of capital sentencing procedures is a disincentive to seeking the death penalty, and a legislature concerned to compromise between advocates and opponents of capital punishment could rationally conclude that providing an alternative, less expensive procedure that offered adequate protection in the form of an incapacitative sentence of life without parole would provide an incentive to prosecutors not to seek executions in any but the most extreme cases. That inducement would be greatly reduced if the alternative carried the same costs as the capital procedure itself. To see this as a matter of simple "economizing” at the expense of fair procedure is an oversimplification of complex legislative choices.

16 . All other members of the initial panel had by this time been excused, either for cause, on consent, or by defense peremptory challenge. Although the judge did not make an unambiguous finding (and none was requested by the defense), the court appears to have agreed with the prosecutor that the only juror thus far seated was African-American. Voir Dire Tr. 230-31.

17 . Juror Pettigrew had stated on voir dire that “I could be [impartial and have an open mind] but ... [i]t would be very difficult” given the nature of the evidence that counsel had indicated would be admitted at trial. Voir Dire Tr. 208-09. He also worked nights, and would not be excused from work because of jury service. The judge found that his likely difficulty in remaining alert for the trial under those conditions "alone would be sufficient reason, to exercise a peremptory challenge.” Voir Dire Tr. 232-33.

18 . While “exceptional cases” also may exist "in which exorbitant application of a generally sound [state procedural] rule renders the state ground inadequate to stop consideration of a federal question,” the circumstances of this case do not fit within that "limited category.” Lee v. Kemna, 534 U.S. 362, 122 S.Ct. 877, 885, 151 L.Ed.2d 820 (2002).