Concurring.—My opinion prepared for the court addresses only defendant’s contention that his trial counsel, Jefferson M. Parrish, Jr., failed to provide him with effective assistance, and does so because the resolution of that single issue is all that is needed to dispose of this case. I am, of course, in full agreement with the views expressed in that opinion. I write separately, however, because I believe a discussion of three other contentions—in my view, the only other potentially meritorious guilt-phase issues raised by defendant—seems to help reveal the magnitude of Parrish’s incompetence.
I
Defendant contends that in violation of the Wheeler rule the prosecutor, George Kennedy, used his peremptory challenges to strike Hispanics from the jury on the ground of group bias alone, and that as a result the judgment must be reversed. (People v. Wheeler (1978) 22 Cal.3d 258 [148 Cal.Rptr. 890, 583 P.2d 748].) Assuming for the time being that the merits could be reached in the absence of any attempt on the part of Parrish to raise the point below, the contention, as I shall explain, appears to be meritorious.
The facts relevant to this issue are set forth at length in the majority opinion, ante, at pages 180-181.
In Wheeler this court established the law applicable here. There is a rebuttable presumption that a party exercising a peremptory challenge is doing so on a constitutionally permissible ground. If a party believes his opponent is using his challenges improperly, he must timely raise the point and make a prima facie case of such discrimination. He should (1) make as complete a record as is feasible, (2) establish that the persons excluded are members of a cognizable group, and (3) show a strong likelihood that such persons are being challenged because of their group association. To make such a showing, the party may demonstrate that his opponent has struck most or all of the members of the identified group or has used a disproportionate number of his peremptories against the group. He may also show that the jurors in question share only membership in the group and in all other respects are heterogeneous. He may also establish that his opponent failed to engage these jurors in more than desultory voir dire, or indeed to ask them any questions at all. Lastly, he may show that he is himself a member of the excluded group and that his alleged victim is a member of the group to which the majority of the remaining jurors belong. If the court finds that a prima facie case has been made, the burden shifts to the other party to show the peremptory challenges in question were properly exercised. To sustain his burden he must satisfy the court such peremptories were exercised for reasons of specific bias. If the court finds that he does not *229sustain this burden, the presumption of the validity of the questioned peremptory challenges is rebutted. (22 Cal.3d at pp. 280-282.)
The record appears to contain facts sufficient to support an inference that the prosecutor was using his peremptory challenges to strike prospective jurors on the ground of group bias alone. To begin with, Robert A. Bagnod was a member of the Hispanic community, Amelia Maes, Lupe Jiminez, Jose Flores, and C. Carrie Corral evidently were, and Alan F. Porcella may have been. Hispanics are, of course, a cognizable class for purposes of Wheeler analysis. (People v. Trevino (1985) 39 Cal.3d 667, 683 [217 Cal.Rptr. 652, 704 P.2d 719]; People v. Harris (1984) 36 Cal.3d 36, 51 [201 Cal.Rptr. 782, 679 P.2d 433] [plur. opn.].)
Further, there seems a strong likelihood that these prospective jurors were challenged because of their ethnicity rather than because of any specific bias. First, they appear to share only membership in the Hispanic community and to differ from each other in other respects. For example, Bagnod was married, worked in a blue collar job, and had no friends in law enforcement; Maes was 37 years old, was employed by an electronics firm evidently in a managerial or supervisory position, was married to a white collar worker, and had 5 children between the ages of 5 and 17; Jiminez was apparently in her mid to late 20’s, was married to a man who was employed by an automobile manufacturer, had 2 children, and worked outside the home; Flores was an aircraft mechanic, had 3 teenage children, and was married to a woman who worked in a commercial laundry; Corral was a nurse and was apparently near retirement age, her husband was a retired blue collar worker, and they had a 28-year-old married son; Porcella was a community college instructor, had 4 young children, and was married to a woman who did not work outside the home, and had relatives in law enforcement; Michael A. Logan was 20 years old and married.
Second, the circumstance surrounding the exercise of the peremptory challenges are significant. Specifically, the prosecutor struck Bagnod after he engaged in brief and unremarkable voir dire. He struck Maes, Jiminez, Flores, Corral, Porcella, and Logan without conducting any examination at all.
Third, defendant is himself a member of the Hispanic community. The significance of this factor, however, appears to be reduced somewhat by the fact that the victim too was Hispanic.
Finally, the challenged Hispanic prospective jurors “included a number of individuals whose background indicated that, absent their [ethnicity], they would in all probability have been considered desirable jurors by the *230prosecutor . . . .” (,People v. Allen (1979) 23 Cal.3d 286, 294 [152 Cal.Rptr. 454, 590 P.2d 30].) For example, Maes said that a first cousin of hers had been the victim of a homicide and expressed views strongly favorable to the death penalty. Flores also expressed views favorable to the death penalty. Corral stated that about four years previously her home had been burglarized and some items of personal property, including her husband’s credit cards, had been stolen, that a suspect was apprehended but subsequently convicted only of using the credit cards, and that in her opinion the police “did the best they could” on the case.
I would conclude that the foregoing evidence appears to make a prima facie case of group bias.1
The record would not support a finding that the prosecutor was exercising his peremptory challenges properly. As to the challenge to Jiminez, which the prosecutor expressly addressed at the Wheeler hearing, it is doubtful that justification appears. The prosecutor said, “My state of mind on excusing Spanish speaking jurors I did excuse was that, for example, Mrs. Jiminez knew the defendant from high school.” Had Jiminez known or even been acquainted with defendant in high school, the prosecutor’s challenge may perhaps have seemed proper. But on voir dire, Jiminez stated only that she had seen defendant at their high school 13 years before and had had no other contact with him. The Attorney General urges us to speculate that the prosecutor may have acted under the assumption that “a shared experience with the [defendant] could make it difficult for the juror to be as objective as the prosecutor would want.” In view, however, of the minimal nature of the “shared experience” as it appeared under questioning by the court and the defense, and in view of the fact that the prosecutor asked no questions to probe that “experience” or for any other purpose, I decline to indulge in such speculation.
Regarding his challenge of the other prospective jurors, the prosecutor presented no express justification and none clearly appears from the record. The Attorney General suggests some possible reasons for the striking of Maes, Flores, Porcella, Corral, and Logan. But as this court recently emphasized, “[0]ur concern is with the explanation the prosecutor gave to the trial court, not with a theory subsequently devised by the Attorney General for consumption on appeal.” (People v. Turner (1986) 42 Cal.3d 711, 722, *231fn. 7 [230 Cal.Rptr. 656, 726 P.2d 102]; accord, People v. Wheeler, supra, 22 Cal.3d at p. 283, fn. 30.) In any event, the Attorney General frankly admits that Bagnod “presented no apparent reasons for his challenge” and declines. even to attempt to suggest any. In so doing, he effectively concedes that the prosecutor improperly struck Bagnod and thereby violated the Wheeler rule. The exercise of one improper challenge is, of course, sufficient to establish a violation. This is because “a party is constitutionally entitled to a petit jury that is as near an approximation of the ideal cross-section of the community as the process of random draw permits” (People v. Wheeler, supra, 22 Cal.3d at p.277).
If the prosecutor used his peremptory challenges to strike members of a cognizable group solely on the ground of group bias, as the record here appears to show, reversal would be required. (People v. Wheeler, supra, 22 Cal.3d at p. 283, and cases cited.) Why this is so is plain: “The right to a fair and impartial jury is one of the most sacred and important of the guaranties of the constitution. Where it has been infringed, no inquiry as to the sufficiency of the evidence to show guilt is indulged and a conviction by a jury so selected must be set aside.’ ” (Id. at p. 283, quoting People v. Riggins (1910) 159 Cal. 113, 120 [112 P. 862].)
Although defendant’s Wheeler contention thus appears to be substantial, I believe the merits could not be reached: because Parrish made no attempt whatever to raise the issue at any time during the proceedings below, the point is not preserved for appeal but must be deemed waived (People v. Haskett (1982) 30 Cal.3d 841, 860 [180 Cal.Rptr. 640, 640 P.2d 776]; People v. Jurado (1981) 115 Cal.App.3d 470, 491-492, fn. 20 [171 Cal.Rptr. 509]).
In holding that “it is necessary that a Wheeler objection be made at the earliest opportunity during the voir dire process,” the court in People v. Ortega (1984) 156 Cal.App.3d 63, 69 [202 Cal.Rptr. 657], gave “several practical reasons.” “First, an early objection will facilitate the moving party’s counsel in making the best possible prima facie case. Second, an early objection will place the opposing party on notice so that counsel may consider whether and on what basis to continue using peremptories against cognizable group members and to prepare to make the best explanation feasible. Third, an early objection will alert the court so that it can intelligently rule on the questions of prima facie case and, if one is found, explanations. In other words, this procedure will insure that the court will pay close attention to the questions asked of and answered by the jurors and other matters bearing on the use of peremptory challenges. The longer the party waits to make a Wheeler motion the less feasible it will be for the court to recall specific questions and answers and the demeanor of the jurors. *232[([] Fourth, this procedure will promote the efficient and economic administration of justice by permitting the court, if it finds discrimination in the use of peremptory challenges, to dismiss the existing jury panel and obtain a new panel without having to wait until the selection process has been completed, [[[] Finally, this procedure will help the courts and parties achieve the most fair and correct result, both below and on appeal.” (Id. at pp. 69-70.)
Here, because Parrish neither raised the issue nor pressed it after the court had done so, the record is not as complete as a reviewing court would need to determine whether a prima facie case of group bias appears. For example, a court could not discern what percentage of the Hispanic prospective jurors the prosecutor struck or whether he used a disproportionate number of his peremptories against the group.
Moreover, because Parrish did nothing—and the court, evidently as a result of his nonfeasance, did little more—to put the prosecutor on notice that a justification of his use of peremptories was required, the record is not as complete as a reviewing court would need to determine whether the prosecutor struck the prospective jurors on proper grounds.
Hence, because the record is incomplete and the incompleteness must be attributed to the failure of Parrish to raise the Wheeler issue or to press it when the court did so on its own motion, I would conclude the point has not been preserved for appeal and accordingly woulcj be compelled to reject the contention without reaching the merits.
II
Defendant contends that the intercepted telephone call should not have been admitted and that its introduction into evidence was prejudicial. Assuming for the time being that the merits could be reached in the absence of any attempt on the part of Parrish to raise the point below, the contention, as I shall explain, appears to be sound.
The applicable substantive law is well settled. Under the Fourth Amendment a warrantless entry by the police into a residence is presumptively unreasonable and therefore unlawful. (E.g., Welsh v. Wisconsin (1984) 466 U.S. 740, 749-750 [80 L.Ed.2d 732, 742-743, 104 S.Ct. 2091, 2097]; People v. Ramey (1976) 16 Cal.3d 263, 270-276 [127 Cal.Rptr. 629, 545 P.2d 1333].) The rule applies whether the entry is made to search for evidence (e.g., Coolidge v. New Hampshire (1971) 403 U.S. 443, 453-455 [29 L.Ed.2d 564, 575-576, 91 S.Ct. 2022]; People v. Dumas (1973) 9 Cal.3d 871, 882 [109 Cal.Rptr. 304, 512 P.2d 1208]) or to seize a person (e.g., *233Welsh, supra, at p. 748 [80 L.Ed.2d at p. 742, 104 S.Ct. at p. 2097]; Ramey, supra, at pp. 270-276).
When exigent circumstances are present, failure to comply with the warrant requirement is not fatal. (E.g., Welsh, supra, 466 U. S. at p. 748 [80 L.Ed.2d at p. 742, 104 S.Ct. at p. 2097]; Ramey, supra, 16 Cal.3d at pp. 270-277.) It is established, however, that “exceptions to the warrant requirement [under the exigent-circumstances rubric] are “few in number and carefully delineated’. . . .” (Welsh, supra, at p. 749 [80 L.Ed.2d at p. 743, 104 S.Ct. at p. 2097]; see, e.g., Ramey, supra, at pp. 270-277.) Only two exceptions are potentially applicable on the facts before us: hot pursuit of a fleeing felon (United States v. Santana (1976) 427 U.S. 38, 42-43 [49 L.Ed.2d 300, 305-306, 96 S.Ct. 2406]; Warden v. Hayden (1967) 387 U.S. 294, 298-299 [18 L.Ed.2d 782, 787-788, 87 S.Ct. 1642]), and prevention of the imminent destruction or removal of evidence (Vale v. Louisiana (1970) 399 U.S. 30, 35 [26 L.Ed.2d 409, 413-414, 90 S.Ct. 1969]). The United States Supreme Court, moreover, has made it plain that each of these two exceptions must be narrowly construed. The former requires an “immediate or continuous pursuit of the [felon] from the scene of a crime.” (Welsh, supra, at p. 753 [80 L.Ed.2d at p. 745, 104 S.Ct. at p. 2099].) The latter requires that the threatened destruction or removal be imminent. (See Vale, supra, at p. 35 [26 L.Ed.2d at pp. 413-414].)
Failure to comply with the warrant requirement is also not fatal when consent is given. (E.g., Vale, supra, at p. 35; People v. Leib (1976) 16 Cal.3d 869, 873 [129 Cal.Rptr. 433, 548 P.2d 1105].) To be effective, consent must be voluntary. (E.g., Schneckloth v. Bustamonte (1973) 412 U.S. 218, 222 [36 L.Ed.2d 854, 860, 93 S.Ct. 2041]; see, e.g., Bumper v. North Carolina (1968) 391 U.S. 543, 548 [20 L.Ed.2d 797, 802, 88 S.Ct. 1788].)
“Voluntariness is a question of fact to be determined from all the circumstances . . . .” (Schneckloth v. Bustamonte, supra, 412 U.S. at pp. 248-249 [36 L.Ed.2d at p. 875].) Among the factors bearing on the issue is whether “consent has been obtained while the consenting party was confronted by many police officers. The presence of a number of policemen is likely to suggest that the police are contemplating an undertaking which does not depend upon the cooperation of the individual from whom permission to search is being sought. Thus, although it is true that ‘the presence of a large number of officers . . . does not present a situation which is per se coercive,’ that fact in tandem with others may well result in a finding that the consent was not voluntary.” (2 LaFave, Search and Seizure (1978) § 8.2(b), p. 642, fn. omitted [hereinafter LaFave].) Another factor is whether the consenter experienced a “significant interruption of his liberty *234of movement as a result of police actions’ [citation].” (People v. Springer (1983) 92 App.Div.2d 209, 213 [460 N.Y.S.2d 86]; see generally People v. James (1977) 19 Cal.3d 99, 111, fn. 10 [137 Cal.Rptr. 447, 561 P.2d 1135].)
It is important to observe, moreover, that “When the police are relying upon consent as the basis for their warrantless search, they have no more authority than they have been given by the consent.” (2 LaFave, supra, § 8.1(c), p. 624, and authorities cited; accord, People v. Harwood (1977) 74 Cal.App.3d 460, 466 [141 Cal.Rptr. 519]; People v. Superior Court (Arketa) (1970) 10 Cal.App.3d 122, 127 [89 Cal.Rptr. 316].)
If the challenged police conduct is violative of the Fourth Amendment, of course, the exclusionary rule requires that all evidence obtained as a result of such conduct be suppressed. (E.g., Mapp v. Ohio (1961) 367 U.S. 643, 646-660 [6 L.Ed.2d 1081, 1085-1093, 81 S.Ct. 1684]; People v. Cahan (1955) 44 Cal.2d 434, 445-450 [282 P.2d 905, 50 A.L.R.2d 513].)
The law on the evidentiary burdens to which the parties are subject is also well settled. The defendant bears the burden of showing that the police officers acted unlawfully. (Badillo v. Superior Court (1956) 46 Cal.2d 269, 272 [294 P.2d 23].) When the defendant has shown a warrantless entry, the burden shifts to the prosecution to prove that the entry was nevertheless reasonable. (Ibid.; see, e.g., Welsh, supra, 466 U.S. at p. 748 [80 L.Ed.2d at p. 742, 104 S.Ct. at p. 2097]; Ramey, supra, 16 Cal.3d at pp. 270-277.) The prosecution may carry its burden by showing the presence of exigent circumstances—a burden that is heavy indeed. (Welsh, supra, at p. 748 [80 L.Ed.2d at p. 742, 104 S.Ct. at p. 2097]; see, e.g., Ramey, supra, at pp. 270-277.) It may also carry its burden by showing voluntary consent. (E.g., Bumper v. North Carolina, supra, 391 U.S. at p. 548 [20 L.Ed.2d at p. 802]; People v. Shelton (1964) 60 Cal.2d 740, 744-745 [36 Cal.Rptr. 433, 388 P.2d 665].) When the prosecution relies on consent, it must also “prove that [the search] was within the scope of the consent given.” (People v. Harwood, supra, 74 Cal.App.3d at p. 466.)
On my review of the record, the intercepted telephone call appears to have been inadmissible as the product of unlawful police conduct. It is undisputed that Officers Webster, Habina, and Guerra entered defendant’s apartment without a warrant. It does not appear that exigent circumstances existed. The police do not seem to have been engaged either in the “immediate or continuous pursuit of the [felon] from the scene of a crime" (Welsh, supra, 466 U.S. at p. 753 [80 L.Ed.2d at p. 745, 104 S.Ct. at p. 2099]), or in the prevention of the imminent destruction or removal of evidence (Vale, supra, 399 U.S. at p. 35 [26 L.Ed.2d at pp. 413-414]). *235Rather, they seem simply to have been attempting to determine defendant’s whereabouts.
Moreover, it appears doubtful that Lawrence Santiago and Millie Dominguez voluntarily consented to the entry. Although Officer Guerra testified that she and the other officers were “invited in,” it seems that on being confronted by three armed and uniformed officers Santiago and Dominguez had little choice in the matter. But even if they voluntarily consented to the entry, there is no evidence that they gave consent to search. And in any event they do not appear to have consented to the interception of the telephone call. On their arrival, the police announced they were looking for defendant. Hence, any consent Santiago and Dominguez gave on the officers’ arrival appears to have extended only to a search for defendant’s person. Once the police entered, they ordered Santiago and Dominguez to remain in the living room and not to move, then searched for defendant throughout the apartment. After they finished their search without finding defendant, the telephone rang; the police prohibited Santiago and Dominguez from answering it; and Officer Guerra picked up the receiver, identified herself in Spanish as “Millie,” and intercepted the statement allegedly made by defendant. In view of the fact that at the time the telephone rang Santiago and Dominguez were “confronted by many police officers” (2 LaFave, supra, § 8.2(b), p. 642), and had experienced a “significant interruption of [their] liberty of movement as a result of police action’ ” (People v. Springer, supra, 92 App.Div.2d at p. 213), they do not appear to have given voluntary consent to the interception of the telephone call.
The introduction of the intercepted telephone call could not easily be held harmless. Because defendant’s right to be free of unreasonable searches and seizures guaranteed by the Fourth Amendment is involved, the court would be compelled to apply the reversible error test set out in Chapman v. California (1967) 386 U.S. 18 [17 L.Ed.2d 705, 87 S.Ct. 824, 24 A.L.R.3d 1065]. That test provides that “before a federal constitutional error can be held harmless, the court must be able to declare a belief that it was harmless beyond a reasonable doubt.” (Id. at p. 24 [17 L.Ed.2d at pp. 710-711].) The Chapman court reiterated the approach it had adopted in Fahy v. Connecticut (1963) 375 U.S. 85 [11 L.Ed.2d 171, 84 S.Ct. 229]: “The question is whether there is a reasonable possibility that the evidence complained of might have contributed to the conviction.” (Id. at pp. 86-87 [11 L.Ed.2d at p. 173].) When a fundamental constitutional right is at issue, under this standard erroneous evidentiary rulings are seldom harmless: “An error in admitting plainly relevant evidence which possibly influenced the jury adversely to a litigant cannot... be conceived of as harmless.” (Chapman v. California, supra, at pp. 23-24 [17 L.Ed.2d at pp. 710-711].) Here the intercepted telephone call was plainly relevant to the case, tending to show *236as it does that defendant was involved in the August 1978 robbery and consequently had a motive to commit murder. Moreover, because the statement ment was allegedly made by defendant himself, the intercepted telephone call may well have influenced the jury.
Even though the contention appears to be substantial, I believe the merits could not be reached because of Parrish’s failure to preserve the issue by a timely challenge of the police conduct. (People v. Lilienthal (1978) 22 Cal.3d 891, 896 [150 Cal.Rptr. 910, 587 P.2d 706]; People v. Jenkins (1975) 13 Cal.3d 749, 753 [119 Cal.Rptr. 705, 532 P.2d 857]; People v. Gallegos (1971) 4 Cal.3d 242, 249 [93 Cal.Rptr. 229, 481 P.2d 237].) Had Parrish made such a challenge, the prosecutor would have been put on notice and might have presented evidence and argument to show that the entry was lawful or that the intercepted telephone call was admissible. (People v. Flores (1968) 68 Cal.2d 563, 567 [68 Cal.Rptr. 161, 440 P.2d 233].) Hence, I would conclude the point has not been preserved for appeal and accordingly would be compelled to reject the contention without reaching the merits.
Ill
Defendant contends that the prosecutor engaged in misconduct on several occasions. Assuming for the time being that the point is preserved on appeal in spite of the fact that Parrish failed in each instance to make any objection whatever, the contention appears to have merit.
The term “prosecutorial misconduct” implies “a dishonest act or an attempt ... to persuade the court or jury by the use of deceptive or reprehensible methods.” (People v. Beivelman (1968) 70 Cal.2d 60, 75 [73 Cal.Rptr. 521, 447 P.2d 913]; accord, People v. Chojnacky (1973) 8 Cal.3d 759, 766 [106 Cal.Rptr. 106, 505 P.2d 530] [plur. opn.].) Accordingly, the term includes all acts designed to obtain victory by improper means. “Whether the prosecution has committed misconduct depends upon the particular circumstances of each case and bad faith must be shown to establish the existence of misconduct . . . .” (People v. Romo (1975) 47 Cal.App.3d 976, 987 [121 Cal.Rptr. 684]; accord, People v. Gomez (1976) 63 Cal.App.3d 328, 338 [133 Cal.Rptr. 731].)
Defendant raises several complaints. First, he claims the prosecutor engaged in misconduct by striking Hispanics from the jury on the ground of group bias alone. Because the Wheeler issue, as I have explained, appears to have merit, so too does this point.
Defendant also claims that the prosecutor’s repeated comments and questions relating to the victim’s extrajudicial identification were deliberate at*237tempts to place before the jury evidence he had committed himself not to introduce, and as such amounted to misconduct. This point also appears to have merit.
Although the prosecutor refrained from introducing the extrajudicial identification directly, he nevertheless repeatedly placed it before the jury through comments and questions. For example, at the beginning of his opening statement he made his first reference to the identification: “Mr. Flores got a license number that was involved and gave the license number to the police. The police did some investigating with respect to the license number. Then they compiled a folder with a lot of photographs in the folder. One of these photographs was of Mr. Ledesma. The San Jose Police Department then took this folder with the photographs, and including photographs of Mr. Ledesma, out to the gas station and they contacted Mr. Flores there and asked him if he recognized anyone in these photographs, in no way indicating that one of these people was a suspect, [fl] Now, as a result of that, a complaint was issued against Mr. Ledesma being identified as the man who committed the robbery of the gas station ... .” (Italics added.) At another point in his opening statement he said: “And further [defendant] was quite upset with Mr. Flores, hard-working young man, for identifying him, having the nerve to identify Mr. Ledesma as the robber.” (Italics added.)
In his questioning of witnesses the prosecutor unmistakably implied that the victim had identified defendant. Examining Sylvia Ontiveros, he asked: “Did Fermín ever tell you how he found out that the witness identified him?”
In his closing argument the prosecutor placed the extrajudicial identification before the jury yet again. At one point he said: “You have a motive, you have flight, you have the license number, you have the identification from a photograph.” (Italics added.) At another he stated: “It isn’t just the license number. It’s the motive. It’s the identification.” (Italics added.) At still another point he said: “[If you vote for acquittal,] You’re going to be saying that it’s reasonable that there was a misidentification by the victim when the victim was just looking at photographs, didn’t have any idea about the size, the way the person talked. Didn’t know which one had the motorcycle when he was looking at the pictures and identified Mr. Ledesma. All he had was a face. Picked him out, just looking at the face. Not knowing anything else about him, not hearing his speech, not knowing that he had the motorcycle that was used in the crime. Not knowing how the man was dressed and so on. Just a face. That’s the man he picked out. And saying all of these things could have been cooked up, could have been contrived.”
*238It is misconduct for the prosecutor to refer in his argument to facts not in evidence. (E.g., People v. Kirkes (1952) 39 Cal.2d 719, 724 [249 P.2d 1]; People v. Evans (1952) 39 Cal.2d 242, 251 [246 P.2d 636].) It is also misconduct for him to suggest the existence of such facts through the questions he asks. (People v. Wagner (1975) 13 Cal.3d 612, 619-620 [119 Cal.Rptr. 457, 532 P.2d 105], and cases cited.)
Under these rules the prosecutor plainly appears to have engaged in misconduct: in argument and examination he repeatedly referred to the victim’s extrajudicial identification of defendant and treated it as though it were a fact in evidence. The prosecutor’s comments and questions were improper because through them he manifestly intended to do, and did in fact do, indirectly the very thing that he had committed himself not to do directly— i.e., place before the jury the victim’s extrajudicial identification of defendant.
I recognize, to be sure, that jurors are commonly instructed—as were the jurors here—that statements by counsel are not evidence. I also recognize that it must generally be presumed that jurors can and will follow the instructions given to them. Nevertheless, I do not believe that the instruction that counsel’s statements are not evidence makes the comments and questions at issue here proper or even cured them of their taint. The danger that such comments and questions may mislead is by no means imaginary or negligible. Indeed, it is plain that the jurors in this case were in fact misled. Although in his testimony about the photographic display Sergeant Traskowski did not actually say that the victim had identified defendant, and although the prosecutor emphasized at one point in closing argument that Traskowski did not so testify, the jurors evidently believed that such testimony had indeed been given: during deliberations their first request to the court was for a rereading of Traskowski’s testimony “re: positive I.D., i.e., I.D. of Ledesma.”
The Attorney General argues in effect that the prosecutor’s comments and questions were proper because they merely stated or implied what the jurors could reasonably infer from the evidence, i.e., that the victim identified defendant. The argument must be rejected. Although a prosecutor may generally present the jurors with conclusions fairly inferable from the evidence, he may not do so when as here the conclusion effectively places before the jurors evidence he has committed himself not to introduce: otherwise, as we have explained, he would be allowed to do indirectly what he has in effect prohibited himself to do directly.
Defendant also complains that the prosecutor improperly elicited testimony that anonymous telephone callers told the police that he was the *239perpetrator of crimes charged. The point is supported by the following portion of the prosecutor’s cross-examination of defendant.
“Q. You have read all the police reports, haven’t you? A. Yes, I have.
“Q. And do you have any idea why all those people phoned in, saying T don’t want my name known, but’— A. Yeah, you—
“Q. ‘Fermín is the guy that did it.’ A. What?
<6
“Q. (By Mr. Kennedy) You were just talking about the police reports a few minutes ago. Do you remember that? A. Um-hum. I have gone over them a couple of times.
“Q. And you indicated that you knew that a bunch of people phoned in and talked about how it happened and who did it and so forth? A. Um-hum.
“Q. Now, why do you suppose all those people wanted to frame you up, back there at that time? A. I don’t know. I don’t even know some of the people. Some people said Jack in the Box—they heard it from other people, you know, some lady says that some guy works with her. He’s just got out of prison or somebody is on parole. You don’t want his name mentioned and—so he’s going through her.
“Q. All right. Let’s talk about her for a minute. This is the lady who works at a local factory? A. I have to look at my notes. Can I get them?
“Q. Well, let’s just talk about what you remember, if you don’t mind.
“Q. (By Mr. Kennedy) Now, you indicated that one of these anonymous tips that you recall was from a man who overheard you talking at the Jack in the Box. Isn’t that right? A. I didn’t say that.”
Also relevant on the point is the question the prosecutor posed to Sergeant Kahn on redirect examination on rebuttal and the answer the officer gave: “Q. Have [the anonymous tipsters] all said Fermín was the one who actually did the killing . . .? A. That’s correct.”
It is misconduct for a prosecutor to “intentionally elicit inadmissible testimony.” (People v. Sims (1976) 64 Cal.App.3d 544, 554 [134 Cal. Rptr. *240566]; accord, People v. King (1968) 266 Cal.App.2d 437, 464 [72 Cal.Rptr. 478], citing authorities.) Here it would be difficult to find that the prosecutor did not know that the evidence he elicited was inadmissible—it was plain hearsay—or that he did not intentionally elicit such evidence—the purpose of his questioning was manifestly to obtain the very evidence he did in fact obtain. Thus, in this regard the prosecutor appears to have engaged in misconduct.
Turning from the question whether the prosecutor engaged in misconduct, I observe that it would be difficult to hold that the misconduct was harmless. The case was evidently close; therefore, the extrajudicial identification was crucial in establishing who robbed the victim on August 26, 1978, and hence who killed him a week and a half later. It follows that there would be a reasonable probability that but for the prosecutor’s misconduct an outcome more favorable to defendant would have resulted.
Nevertheless, although defendant’s claim that the prosecutor engaged in prejudicial misconduct appears substantial, I believe it is not preserved for appeal and hence could not be resolved on its merits but would have to be rejected at the threshold. To preserve such a point, “miscondúct must be assigned as error at trial with a request that the jury be instructed to disregard its effect . . . .” (People v. Chi Ko Wong (1976) 18 Cal.3d 698, 723 [135 Cal.Rptr. 392, 557 P.2d 976], disapproved on another ground in People v. Green (1980) 27 Cal.3d 1, 34 [164 Cal. Rptr. 1, 609 P.2d 468].) Simply to object or make the assignment of misconduct without seeking a curative admonition is generally not enough. (People v. Beivelman, supra, 70 Cal.2d at p. 75.) “The reason for this rule, of course, is that ‘the trial court should be given an opportunity to correct the abuse and thus, if possible, prevent by suitable instructions the harmful effect upon the minds of the jury.’ ” (People v. Green, supra, at p. 27.) Here Parrish failed to make any objection whatever to any of the instances of misconduct raised on appeal.
To the rule requiring an assignment of misconduct and a request for a curative admonition in order to preserve a claim of prosecutorial misconduct, there is, as the court held in People v. Green, supra, 27 Cal.3d at page 34, a single exception: a reviewing court may nevertheless reach the merits of such a claim if a timely assignment of misconduct and request for admonition would not have cured the harm. The exception, however, appears not to be applicable on this record: defendant fails to show that any instance of prejudicial misconduct was incurable. This is clear, for example, with regard to the prosecutor’s references to the victim’s extrajudicial identification of defendant. A timely assignment of misconduct and admonition to the first comment, which seems not prejudicial in itself, would likely *241have cured the harm and would also, I assume, have deterred the prosecutor from engaging in further misconduct of the same sort. “There is accordingly no ground to excuse defendant from the general requirement of a timely objection, and the point must be deemed waived.” (People v. Green, supra, at p. 35.)
Defendant argues that prior to Green a second exception existed—“where the case is closely balanced and there is grave doubt of defendant’s guilt, and the acts of misconduct are such as to contribute materially to the verdict” (People v. Berryman (1936) 6 Cal.2d 331, 337 [57 P.2d 136])—and that because Green was decided after his trial was completed, this exception should remain available to him.
It is of course the general rule that “judicial decisions apply ‘retroactively.’ [Citation.] Indeed, a legal system based on precedent has a built-in presumption of retroactivity.” (Solem v. Stumes (1984) 465 U.S. 638, 642 [79 L.Ed.2d 579, 586, 104 S.Ct. 1338, 1341]; accord, People v. Guerra (1984) 37 Cal.3d 385, 399 [208 Cal.Rptr. 162, 690 P.2d 635].) In reliance on People v. Guerra, supra, at pages 399-402, defendant maintains that Green should be given only prospective effect because it established a new rule of law that was contrary to a prior rule. I am not persuaded.
Green merely restated the well-settled rule that the issue of prosecutorial misconduct is preserved for appeal only if a timely objection was made. That decision did indeed disapprove the language in Berryman quoted above. But that language, which had neither historical legitimacy nor analytical soundness, had been “ignored” by “a parallel line of cases in this court” (People v. Green, supra, 27 Cal.3d at p. 34), and hence cannot be characterized as a prior contrary rule on which justifiable reliance could have been placed and which would thereby bar the normal retroactive operation of Green.
Even if Green had in fact established a new rule when there was a prior contrary rule, prospectivity would not be compelled. In Guerra the court clearly implied that except where the purpose of the new rule points plainly toward prospectivity, the existence of “justifiable reliance on an old rule to the contrary” is necessary before “the courts may choose to make, on grounds of policy, an exception to ‘the ordinary assumption of retrospective operation’ [citation].” (37 Cal.3d at p. 401.) At the threshold, the purpose of the Green rule does not appear to point plainly toward prospectivity, nor does defendant contend otherwise. The possibility of prospectivity, therefore, turns on whether there was justifiable reliance. Plainly there was none: *242no reasonably competent defense counsel would have relied on the “exception” stated in Berryman to preserve the issue of prosecutorial misconduct for appeal.
Accordingly, because I believe the point does not survive for purposes of appellate review, I would be compelled to reject defendant’s claim of prosecutorial misconduct without reaching its merits.
In sum, Parrish’s incompetence was of such magnitude that it not only completely undermined the reliability of the trial proceedings but also threatened to preclude defendant from raising before this court the only contentions—other than, of course, the ineffective-assistance claim—that were potentially meritorious on the issue of guilt. In light of such incompetence and the resulting prejudice to which it subjected the defense, the conclusion is inescapable: Parrish failed to provide defendant with the effective assistance of counsel.
I recognize that my conclusion is contrary to that of the trial court. I commend the court for raising the Wheeler issue sua sponte. But for the foregoing reasons I am compelled to disagree. Fault, however, should be placed not with the court but with Parrish: it was his responsibility to assist the court by at least attempting to make a prima facie case; but in the face of significant evidence that the prosecutor was unconstitutionally exercising his peremptory challenges, he simply did nothing.