Docket No. Crim. 9514.
February 8, 1967.
APPEAL from a judgment of the Superior Court of Los Angeles County. Alfred P. Peracca, Judge. Reversed.
Howard E. Beckler, under appointment by the Supreme Court, for Defendant and Appellant.
Thomas C. Lynch, Attorney General, William E. James, Assistant Attorney General, and George J. Roth, Deputy Attorney General, for Plaintiff and Respondent.
The present appeal requires us to determine once again which cases are to be governed in this state by the principles set forth in Escobedo v. Illinois (1964) 378 U.S. 478 [12 L.Ed.2d 977, 84 S.Ct. 1758], and to decide for the first time which cases, among those governed by Escobedo, should also be subject to the more detailed guidelines enunciated in Miranda v. Arizona (1966) 384 U.S. 436 [16 L.Ed.2d 694, 86 S.Ct. 1602, 10 A.L.R.3d 974].
As we point out in greater detail hereinafter, the matter is not one of first impression in California.
This question has not previously been presented by any of our cases.
The trial in this case began on May 13, 1963. Defendant, George Fulton Rollins, and his codefendants, Willie Slater and John Palmer, were charged with two counts of armed robbery. All three defendants waived the right to a jury trial, and the trial court found each defendant guilty on both counts. The court further found that Slater and Palmer committed the robberies while armed with a deadly weapon but that Rollins was not personally armed. On June 7, 1963, the court entered the judgment which defendant Rollins challenges in this appeal, convicting him on two counts of first degree robbery. (Pen. Code, §§ 211, 211a.) On June 20, 1963, the court imposed concurrent sentences and committed Rollins to the California Youth Authority for the term prescribed by law.
On June 22, 1964, the United States Supreme Court decided Escobedo v. Illinois, supra, 378 U.S. 478; on June 13, 1966, while this appeal was still pending, the United States Supreme Court filed its decision in Miranda v. Arizona, supra, 384 U.S. 436. In this setting, we must first decide what law governs; we must then apply that law to the judgment here on review.
Under Johnson v. New Jersey (1966) 384 U.S. 719 [16 L.Ed.2d 882, 86 S.Ct. 1772], the states are not bound to apply either Escobedo or Miranda to trials commencing prior to the dates when those decisions were rendered. Under our ruling in In re Lopez (1965) 62 Cal.2d 368 [ 42 Cal.Rptr. 188, 398 P.2d 380], however, the principles established in Escobedo v. Illinois, supra, 378 U.S. 478, and People v. Dorado (1965) 62 Cal.2d 338 [ 42 Cal.Rptr. 169, 398 P.2d 361], are applicable to the instant case since the judgment of conviction was not final before June 22, 1964, the date of the Escobedo decision.
We have decided that we should follow Johnson insofar as that decision limited the operation of Miranda in this state to trials beginning after June 13, 1966. In accordance with this determination of the governing law, we have here applied Escobedo and Dorado but not Miranda; we have concluded that the conviction of the defendant Rollins cannot stand.
The two robberies here involved took place in service stations in Los Angeles. James Carter, an attendant at the U-Save-On service station, testified at the trial that at 12:50 on the morning of January 28, 1963, two men whom he identified as codefendants Slater and Palmer walked into the station and held him up at gunpoint. They took bills, change, a coin tray, and cigarettes, and then ran off.
Arthur Hammond, an attendant at the Jet 777 service station, testified that at 1:30 on the same morning Slater and Palmer came into the station and held him up at gunpoint. They took bills, change, cigarettes, a coin changer of the type used by a bus driver, and Hammond's wallet.
Officer Helvin of the Los Angeles Police Department testified that he and his partner, Officer Matlock, while on patrol early in the morning of January 28, observed the driver of a black Chevrolet look at their police car and then suddenly make a right turn. The officers followed the Chevrolet for several blocks at speeds up to 60 miles an hour. At one point during the chase the officers noticed an arm extending from the right window of the car. The Chevrolet ran through two stop signs; shortly thereafter the officers stopped the vehicle and asked the driver, defendant Rollins, to step out. Slater and Palmer were also in the car.
Officer Helvin handcuffed Rollins, placed him under arrest, and put him in the patrol car. Suspecting that the vehicle might have been stolen, the officer reached for the registration slip. While doing so, he noticed under the front seat a coin changer of the type used by bus drivers. At approximately this time Officer Helvin heard a report over his police radio about the holdup at the Jet 777 station and learned that a coin changer was among the items taken. He then questioned defendant regarding the coin changer found in the car and as to the robbery; defendant denied any knowledge of either. Defendant said that on his return from Compton, at a point near the Jet 777 station, he had picked up Slater and Palmer, whom he had not previously known.
The officers then retraced the route over which they had pursued defendant; at different points along the route they found Hammond's wallet, a .22 caliber revolver, and a .38 caliber revolver.
Sergeant Chiquet testified that, during the morning of January 29, he and three other officers questioned defendant at the police station. Defendant stated that Slater and Palmer had come to his house about 12:30 on the morning of January 28 and had asked that he drive them somewhere. He said that he had driven them to a point near the Jet 777 station and that they had left him there with the car but had returned several minutes later. Defendant again denied knowledge of any robberies.
At the police station in the afternoon of January 29, Sergeant Chiquet and the three other officers again questioned defendant, this time together with Slater and Palmer. Sergeant Chiquet testified that, at some point during this tape-recorded interrogation, defendant Rollins admitted that he was acquainted with defendants Palmer and Slater and finally confessed, stating that he had driven the car in the service station robberies.
Defendant Rollins took the stand in his own behalf. He testified that he had seen Palmer in a bar shortly before midnight on January 27 and had lent Palmer his car keys. Palmer left the bar and later returned, whereupon defendant and Palmer departed together in defendant's car. At a point near the Jet 777 station Palmer asked defendant, who was driving, to stop the car; defendant did so and Palmer stepped out. When he returned a few minutes later, Slater accompanied him. Defendant then drove away. He saw a police car; when he realized that it was following him, he increased his speed. When the police turned on the siren, he stopped.
Defendant denied any knowledge of the robberies. He admitted that he had told the police that he drove the car involved in the robberies, but he testified that he had made this statement only because he had previously been physically abused by the police. Slater and Palmer also took the stand. Slater denied any participation in the robberies; Palmer admitted that he had participated, but testified that defendant was in no way involved.
Before we test the foregoing record for reversible error, we confront the threshold problem of determining the applicable law in cases tried before the date of the Escobedo decision but not yet final either on that date or on the date of the decision in Miranda. In holding Escobedo and Dorado unavailable on collateral attack to challenge convictions which had become final before the Escobedo decision ( In re Lopez, supra, 62 Cal.2d 368), we reasoned that the primary purpose of Escobedo and Dorado was to prevent police tactics which invited coerced confessions; that past violations of the rules established by those decisions did not affect the fairness of convictions already final so long as truly coerced confessions could be exposed at trial, on appeal, or on collateral attack; and that completely retroactive application of Escobedo and Dorado would impose an undue burden upon the administration of criminal justice. Accordingly, we concluded that Escobedo and Dorado should operate only "prospectively and retroactively to cases on appeal [as of June 22, 1964]. . . ." ( In re Lopez, supra, 62 Cal.2d 368, 378 fn. 16.)
In Johnson v. New Jersey, supra, 384 U.S. 719, 726-734, the United States Supreme Court analyzed the problem of the retroactivity of Escobedo and Miranda in terms similar to those of Lopez. Although our analysis led us to apply Escobedo and Dorado to all convictions which had not become final before June 22, 1964, the United States Supreme Court decided to require the application of Escobedo and Miranda only in trials which began after the respective dates of those decisions. The court was careful, however, to leave the states free to apply Escobedo and Miranda to a broader range of cases. The opinion said: "We recognize that certain state courts have perceived the implications of Escobedo and have therefore anticipated our holding in Miranda. Of course, States are still entirely free to effectuate under their own law stricter standards than those we have laid down and to apply those standards in a broader range of cases than is required by this decision." ( Johnson v. New Jersey, supra, 384 U.S. 719, 733.) While leaving the states free to apply Escobedo and Miranda more broadly, the United States Supreme Court found no "persuasive reason to extend Escobedo and Miranda to cases tried before those decisions were announced, even though the cases may still be on direct appeal." ( Ibid.)
See Mishkin, The Supreme Court, 1964 Term — Foreword: The High Court, The Great Writ, and the Due Process of Time and Law (1965) 79 Harv.L.Rev. 56, 95-101.
As a matter of normal judicial operation, even a non-retroactive decision ordinarily governs all cases still pending on direct review when the decision is rendered. (Compare, e.g., People v. Cahan (1955) 44 Cal.2d 434 [ 282 P.2d 905, 50 A.L.R.2d 513] (Traynor, J.), with In re Harris (1961) 56 Cal.2d 879, 881-886 [ 16 Cal.Rptr. 889, 366 P.2d 305] (Traynor, J., concurring); see also Tehan v. Shott (1966) 382 U.S. 406 [15 L.Ed.2d 453, 86 S.Ct. 459]; Linkletter v. Walker (1965) 381 U.S. 618 [14 L.Ed.2d 601, 85 S.Ct. 1731]; Mishkin, op. cit. supra, 79 Harv.L.Rev. 56, 72, 77-79.)
We share the conclusion of the United States Supreme Court that Miranda should not extend to trials which began before June 13, 1966. The purposes of Miranda point to purely prospective application; collateral relief remains available on proper showing to sift out confessions which were not truly voluntary; retroactive application, requiring retrials in countless cases, would needlessly burden the administration of justice.
Moreover, the application of Miranda to trials beginning before June 13, 1966, would necessarily entail the reversal of some convictions in cases which came to trial after Escobedo and Dorado, and which complied fully with the rules announced in those decisions. The application of Miranda to reverse such convictions would upset the justifiable reliance of those who sought scrupulously to follow the rulings of this court and of the United States Supreme Court.
Escobedo was the first of a series of decisions "to condemn an entire process of in-custody interrogation. . . ." ( Johnson v. New Jersey, supra, 384 U.S. 719, 731.) Law enforcement agencies were confronted for the first time with constitutional imperatives which addressed themselves specifically to the steps which police must take to safeguard the rights to silence and to counsel. In response to the guidelines which Escobedo announced and which Dorado applied, law enforcement authorities "adopted devices which, although below the constitutional minimum [as defined by Miranda], were not intentional evasions of the requirements of the privilege [against self-incrimination]." ( Johnson v. New Jersey, supra, 384 U.S. 719, 733.)
We fully appreciate the difficulties which beset law enforcement agencies in this endeavor: we would compound those difficulties needlessly if we were to require retrials of pre- Miranda cases in which prosecutors and trial courts carefully heeded the teachings of Escobedo and of Dorado but failed to anticipate the additional requirements set forth in Miranda. Under these circumstances, a due regard for the relationship between those who interpret the commands of the Constitution and those who must comply with such interpretations while combatting crime impels the limitation of Miranda to trials commencing after June 13, 1966, the date on which Miranda was decided.
As at least one commentator had observed prior to Johnson, considerations of official reliance point more readily to the date of trial than to the date of final judgment. (Mishkin, op. cit. supra, 79 Harv.L.Rev. 56, 74.)
Although we must respect the reliance evoked by Escobedo and Dorado in cases tried during the period between those decisions and the decision in Miranda, we find little force in the reliance argument as applied to trials which took place before Escobedo was decided on June 22, 1964, since pre- Escobedo law did not encourage the sort of reliance which the Escobedo and Dorado decisions clearly invited.
Prior to Escobedo, the United States Supreme Court had not imposed constitutional limitations directly upon the process of custodial interrogation; instead, it had undertaken to scrutinize the record of each defendant's interrogation in order to decide, under all the circumstances of each case, whether a challenged confession could be deemed truly voluntary. Since the relevant inquiry drew so heavily upon a highly particularized sort of hindsight, and since the outcome of any given case was so difficult to predict with any confidence, police officers and prosecutors can hardly have placed affirmative reliance upon the case law preceding Escobedo in the planning of custodial interrogation or in the preparation and presentation of cases for trial. Thus the application of Escobedo and Dorado to cases tried prior to the date of the Escobedo decision does not, in any meaningful sense, upset official efforts to obey constitutional imperatives; as to such pre- Escobedo trials, we see no reason to depart from the usual rule that even a non-retroactive judicial decision governs all cases which were not yet final when the decision was announced.
See, generally, Developments in the Law, Confessions, (1966) 79 Harv. L.Rev. 935, 954-984; cf. Crooker v. California (1958) 357 U.S. 433 [2 L.Ed.2d 1448, 78 S.Ct. 1287]; Cicenia v. Lagay (1958) 357 U.S. 504 [2 L.Ed.2d 1523, 78 S.Ct. 1297].
See Graham, What is "Custodial Interrogation"?: California's Anticipatory Application of Miranda v. Arizona (1966) 14 U.C.L.A. L.Rev. 58, 89 and fn. 173.
See footnote 3, supra.
The Supreme Court of Rhode Island has adopted a rather unusual approach to the retroactivity problem. In State v. Mendes (1965) ___ R.I. ___ [ 210 A.2d 50], the Rhode Island court concluded, as we had concluded in Dorado, that the rights established by Escobedo could not be limited to suspects who had specifically requested counsel. The Rhode Island Supreme Court recently decided, however, that it would apply Escobedo as interpreted in Mendes only to defendants whose interrogations took place at least 30 days after the date of the Mendes decision on May 10, 1965. ( State v. Gannites (1966) ___ R.I. ___ [ 221 A.2d 620, 623-624].) Still more recently, that court substituted date of arrest for date of interrogation in its retroactivity formula. ( State v. McParlin (1966) ___ R.I. ___ [ 221 A.2d 790, 793].)
Entirely apart from its complexity, the Gannites-McParlin solution to the problem of retroactivity has little to commend it. Even under the rule set forth by the United States Supreme Court in Johnson, the fact that a suspect was arrested or interrogated before a given date is immaterial: Until the trial has begun, it is not unduly burdensome to require the prosecution to forego the use of evidence which fails to meet current constitutional standards.
Moreover, any suggestion that Escobedo should be more broadly available to defendants who requested counsel than to those who did not would, as we pointed out in Dorado, discriminate against those who did not know their rights. ( People v. Dorado, supra, 62 Cal.2d 338, 351; see Escobedo v. Illinois, supra, 378 U.S. 478, 490.)
We note, finally, that this matter is not one of first impression in California; we do not write upon a clean slate, since we have already applied Escobedo to reverse convictions in numerous cases which went to trial before June 22, 1964. Thus we cannot say here, as could the United States Supreme Court in Johnson v. New Jersey, supra, 384 U.S. 719, 832, that "the possibility of applying [ Escobedo] only prospectively is yet an open issue."
See, e.g., People v. Treloar (1966) 64 Cal.2d 141 [ 49 Cal.Rptr. 100, 410 P.2d 620]; People v. Brooks (1966) 64 Cal.2d 130 [ 48 Cal.Rptr. 879, 410 P.2d 383]; People v. Buchanan (1966) 63 Cal.2d 880 [ 48 Cal.Rptr. 733, 409 P.2d 957]; People v. Ketchel (1966) 63 Cal.2d 859 [ 48 Cal.Rptr. 614, 409 P.2d 694]; People v. Smith (1966) 63 Cal.2d 779 [ 48 Cal.Rptr. 382, 409 P.2d 222]; People v. Chaney (1965) 63 Cal.2d 767 [ 48 Cal.Rptr. 188, 408 P.2d 964]; People v. Gilbert (1965) 63 Cal.2d 690 [ 47 Cal.Rptr. 909, 408 P.2d 365]; In re Varnum (1965) 63 Cal.2d 629 [ 47 Cal.Rptr. 769, 408 P.2d 97]; People v. Grubb (1965) 63 Cal.2d 614 [ 47 Cal.Rptr. 772, 408 P.2d 100]; People v. Marbury (1965) 63 Cal.2d 574 [ 47 Cal.Rptr. 491, 407 P.2d 667]; People v. Arguello (1965) 63 Cal.2d 566 [ 47 Cal.Rptr. 485, 407 P.2d 661]; People v. Green (1965) 63 Cal.2d 561 [ 47 Cal.Rptr. 477, 407 P.2d 653]; People v. Faris (1965) 63 Cal.2d 541 [ 47 Cal.Rptr. 370, 407 P.2d 282]; People v. Aranda (1965) 63 Cal.2d 518 [ 47 Cal.Rptr. 353, 407 P.2d 265]; People v. Furnish (1965) 63 Cal.2d 511 [ 47 Cal.Rptr. 387, 407 P.2d 299]; People v. Stockman (1965) 63 Cal.2d 494 [ 47 Cal.Rptr. 365, 407 P.2d 277]; People v. Luker (1965) 63 Cal.2d 464 [ 47 Cal.Rptr. 209, 407 P.2d 9]; People v. Polk (1965) 63 Cal.2d 443 [ 47 Cal.Rptr. 1, 406 P.2d 641]; People v. Price (1965) 63 Cal.2d 370 [ 46 Cal.Rptr. 775, 406 P.2d 55]; People v. Anderson (1965) 63 Cal.2d 351 [ 46 Cal.Rptr. 763, 406 P.2d 43]; People v. Roberts (1965) 63 Cal.2d 84 [ 45 Cal.Rptr. 155, 403 P.2d 411]; People v. Holford (1965) 63 Cal.2d 74 [ 45 Cal.Rptr. 167, 403 P.2d 423]; People v. Clark (1965) 62 Cal.2d 870 [ 44 Cal.Rptr. 784, 402 P.2d 856]; People v. Forbs (1965) 62 Cal.2d 847 [ 44 Cal.Rptr. 753, 402 P.2d 825]; People v. Bostick (1965) 62 Cal.2d 820 [ 44 Cal.Rptr. 649, 402 P.2d 529]; People v. Davis (1965) 62 Cal.2d 791 [ 44 Cal.Rptr. 454, 402 P.2d 142]; People v. Bilderbach (1965) 62 Cal.2d 757 [ 44 Cal.Rptr. 313, 401 P.2d 921]; People v. Sears (1965) 62 Cal.2d 737 [ 44 Cal.Rptr. 330, 401 P.2d 938]; People v. Schader (1965) 62 Cal.2d 716 [ 44 Cal.Rptr. 193, 401 P.2d 665]; People v. Lilliock (1965) 62 Cal.2d 618 [ 43 Cal.Rptr. 699, 401 P.2d 4]; People v. Stewart (1965) 62 Cal.2d 571 [433 Cal.Rptr. 201, 400 P.2d 97], affd. sub nom. Miranda v. Arizona, supra, 384 U.S. 436, 497-499; People v. Modesto (1965) 62 Cal.2d 436 [ 42 Cal.Rptr. 417, 398 P.2d 753]; People v. Dorado, supra, 62 Cal.2d 338; People v. Sharer (1964) 61 Cal.2d 869 [ 40 Cal.Rptr. 851, 395 P.2d 899].
We have made no effort to list the multitude of pre- Escobedo trials in which the courts of appeal of this state have also reversed convictions in accordance with the mandate of Escobedo and Dorado. (See, e.g., the cases cited in Graham, op. cit. supra, 14 U.C.L.A.L.Rev. 59.)
We should stress here that our many reversals in the cases cited above were rendered in full awareness of the fact that the trials in question had all preceded the date on which Escobedo was decided. Indeed, we have repeatedly relied upon that fact in negating the inference that the defendants in these cases had waived their constitutional rights to silence and to counsel by failing to insist upon them, either during interrogation or at trial. (See, e.g., People v. Treolar, supra, 64 Cal.2d 141, 143-144; People v. Brooks, supra, 64 Cal.2d 130, 135; People v. Green, supra, 63 Cal.2d 561, 564 fn. 3; People v. Aranda, supra, 63 Cal.2d 518, 523; People v. Davis, supra, 62 Cal.2d 791, 796; People v. Lilliock, supra, 62 Cal.2d 618, 621.)
The situation which faces us now is analogous to that which confronted the United States Supreme Court when it decided Linkletter v. Walker, supra, 381 U.S. 618, and again when it decided Tehan v. Shott, supra, 382 U.S. 406. In those cases, the court was required to decide whether to limit Mapp v. Ohio (1961) 367 U.S. 643 [6 L.Ed.2d 1081, 81 S.Ct. 1684, 84 A.L.R.2d 933], and Griffin v. California (1965) 380 U.S. 609 [14 L.Ed.2d 106, 85 S.Ct. 1229], to prospective operation only; the court decided to apply Mapp and Griffin not only prospectively, to trials commencing after the Mapp and Griffin decisions, but also retroactively, to trials which had begun, but had not yet led to final convictions, prior to those decisions.
In Johnson v. New Jersey, supra, 384 U.S. 719, 732, the court explained its resolution of the problem in Linkletter and in Shott by noting that the issue of prospectivity arose in those two cases only after the court had already applied Mapp and Griffin to reverse non-final convictions in several cases tried before Mapp and Griffin had been decided. Accordingly, the court concluded in Johnson that it was no longer free, by the time Linkletter and Shott had reached it, to limit the application of Mapp and Griffin to trials which began after those decisions were announced. Linkletter and Shott were thus "necessarily limited to convictions which had become final by the time Mapp and Griffin were rendered." ( Johnson v. New Jersey, supra, 384 U.S. 719, 732.) To adopt a more stringent measure of prospectivity in Linkletter or in Shott would have been indefensibly arbitrary.
In Linkletter, the court noted Ker v. California (1963) 374 U.S. 23 [10 L.Ed.2d 726, 83 S.Ct. 1623], Fahy v. Connecticut (1963) 375 U.S. 85 [11 L.Ed.2d 171, 84 S.Ct. 229], and Stoner v. California (1964) 376 U.S. 483 [11 L.Ed.2d 856, 84 S.Ct. 889], as examples of decisions in which it had already applied Mapp to cases which were "pending on direct review at the time [ Mapp] was rendered." ( Linkletter v. Walker, supra, 381 U.S. 618, 622 fn. 4.)
Likewise, in Shott, the court noted that there was no "question of the applicability of the Griffin rule to cases still pending on direct review at the time it was announced." ( Tehan v. Shott, supra, 382 U.S. 406, 409 fn. 3.) At that point, the Shott opinion cited O'Connor v. Ohio (1965) 382 U.S. 286 [15 L.Ed.2d 337, 86 S.Ct. 445] (per curiam), a case in which the court had vacated a judgment of the Supreme Court of Ohio and had "remanded . . . for further proceedings in light of Griffin v. California, 380 U.S. 609 [14 L.Ed.2d 106, 85 S.Ct. 1229]."
The court sought to distinguish the situation before it in Johnson by noting that, "apart from the application of the holdings in Escobedo and Miranda to the parties before the Court in those cases, the possibility of applying the decisions only prospectively is yet an open issue." ( Johnson v. New Jersey, supra, 384 U.S. 719, 732.) The court's failure to explain the seeming inconsistency between its affirmance of the convictions in Johnson and its reversal of the convictions in Escobedo and in Miranda has evoked predictable criticism from the commentators (see, e.g., The Supreme Court, 1965 Term (1966) 80 Harv.L.Rev. 91, 140-141), but the supposed inconsistency was more apparent than real, since Johnson, like Linkletter and Shott, arose on collateral attack, whereas Escobedo and Miranda, like Mapp and Griffin, arose on direct appeal. Since the convictions involved in Johnson had become final long before either Escobedo or Miranda had been decided, the result in Johnson would have been the same under either a trial-date cutoff or a final-judgment cutoff.
The same cannot be said, however, of the numerous denials of certiorari announced on the day Johnson was decided, since many of the certiorari petitions denied on June 20, 1966, involved cases tried before the decisions in Escobedo or Miranda but still pending on direct review when those decisions were rendered. In cases such as these, the court admittedly applied a trial-date cutoff which it had not applied either in Escobedo or in Miranda. The resulting inconsistency inheres in any attempt to apply a new rule to the parties in the particular case in which it is announced, while otherwise limiting it to purely prospective operation.
Although this technique has been criticized as productive of "intolerable inequality" (Mishkin, op. cit. supra, 79 Harv.L.Rev. 56, 61 fn. 23; see also Bender, The Retroactive Effect of an Overruling Constitutional Decision: Mapp v. Ohio (1962) 110 U.Pa.L.Rev. 650, 675-676), it might well reflect a justifiable fear that refusing to apply new rules to the cases in which they are established could dangerously limit the incentive to appeal. Moreover, such a refusal might be thought to conflict with the "case or controversy" requirement of article III of the federal Constitution. (See Note, Prospective Overruling and Retroactive Application in the Federal Courts (1962) 71 Yale L.J. 907, 933; contra, see Johnson v. New Jersey, supra, 384 U.S. 719, 733 ; Linkletter v. Walker, supra, 381 U.S. 618, 619-622 fn. 3; but see Mishkin, op. cit. supra, 79 Harv.L.Rev. 56, 59 fn. 13.) If a purely prospective decision is ruled out for reasons such as these, the technique followed by the court in Escobedo, Miranda, and Johnson may represent a satisfactory alternative to the disturbing prospect of reversing all convictions still pending on appeal even when little more than doctrinal consistency counsels such a result, and even when concern for official reliance renders it inadvisable.
Whatever considerations may have led the United States Supreme Court to dispose of the matter as it did, our approach to the problem avoids the difficulties of purely prospective decision-making, preserves the advantages of consistency, and recognizes the importance of respecting official reliance: We apply Miranda in no case tried before that decision was rendered, and we continue to follow Escobedo in the few remaining cases which were pending on direct review when that decision was announced.
Similarly, since we have already applied Escobedo and Dorado to reverse numerous convictions obtained in pre- Escobedo trials, we cannot suddenly reject their application to the few remaining defendants whose cases now pose Escobedo and Dorado issues, and whose convictions were pending on direct review when Escobedo was decided; we cannot abandon principles so clearly embodied in precedent and so firmly established in practice.
Nor will our continued application of Escobedo and Dorado to such defendants impose a major burden on the administration of criminal justice. The vast majority of cases in California which had not become final prior to June 22, 1964, have by this time been disposed of on appeal in accordance with the teaching of Escobedo and Dorado. Accordingly, we need not invite the anomalies and the manifest injustice which the rejection of Lopez, at the virtual end of its natural life, would entail.  Thus, while following Johnson in limiting Miranda to trials which began after June 13, 1966, we adhere to Lopez in applying Escobedo and Dorado to the instant case and to all other cases which had not become final before June 22, 1964.
Indeed, some commentators have advocated giving Escobedo an even broader retroactive reach than was accorded by Lopez. (See, e.g., Comment, Linkletter, Shott, and the Retroactivity Problem in Escobedo (1966) 64 Mich.L.Rev. 832, 842-854.)
Having thus determined the applicable law, we set forth our reasons for concluding that the judgment of conviction rests upon a confession violative of Escobedo and Dorado and therefore cannot stand. We note parenthetically that the Attorney General does not seriously dispute the proposition that, if Escobedo and Dorado are to be applied in the instant case, defendant's conviction must be reversed.
 "[W]hen the officers have arrested the suspect and . . . have undertaken a process of interrogations that lends itself to eliciting incriminating statements, the accusatory or critical stage has been reached and the suspect is entitled to counsel." ( People v. Stewart, supra, 62 Cal.2d 571, 577, affd. sub nom. Miranda v. Arizona, supra, 384 U.S. 436, 497-499; Escobedo v. Illinois, supra, 378 U.S. 478, 490-491; People v. Dorado, supra, 62 Cal.2d 338, 353-354.)
 Here defendant admitted his guilt only after he had been in custody for over 34 hours following his arrest. The record discloses that the police arrested Rollins for robbery; that several officers questioned him, once at the scene of the arrest, and twice at the police station; that he initially denied any involvement in the robberies; but that, in the course of the third interrogation, he finally confessed.
In the absence of evidence to the contrary, we cannot assume that defendant's confession resulted from anything other than "a process of interrogations that lends itself to eliciting incriminating statements." ( Escobedo v. Illinois, supra, 378 U.S. 478, 491; People v. Stockman, supra, 63 Cal.2d 494, 498-499; People v. Luker, supra, 63 Cal.2d 464, 473-474; People v. Dorado, supra, 62 Cal.2d 338, 353.) Accordingly, we conclude that the accusatory stage had been reached and that defendant therefore had a right to remain silent and was entitled to the assistance of counsel. ( Escobedo v. Illinois supra, 378 U.S. 478; People v. Dorado, supra, 62 Cal.2d 338.)
 The record does not affirmatively indicate, however, that defendant was advised of these rights or that he knew of them and knowingly surrendered them. Under these circumstances, "no knowing and intelligent waiver of any constitutional right can be said to have occurred." ( Escobedo v. Illinois, supra, 378 U.S. 478, 490 fn. 14; People v. Stewart, supra, 62 Cal.2d 571, 581, affd. sub nom. Miranda v. Arizona, supra, 384 U.S. 436, 497-499; People v. Dorado, supra, 62 Cal.2d 338, 353.)
 Having concluded that defendant's confession was inadmissible, we must reverse the instant conviction, since the erroneous admission of a confession is necessarily prejudicial. ( People v. Schader, supra, 62 Cal.2d 716, 728-731; People v. Dorado, supra, 62 Cal.2d 338, 356; People v. Parham (1963) 60 Cal.2d 378, 385 [ 33 Cal.Rptr. 497, 384 P.2d 1001].)
Defendant's other extrajudicial statements, as well as his trial testimony, were exculpatory in character; we do not deal here with a case involving multiple confessions. (Cf. People v. Cotter (1965) 63 Cal.2d 386, 398 [ 46 Cal.Rptr. 622, 405 P.2d 862]; People v. Jacobson (1965) 63 Cal.2d 319, 330-331 [ 46 Cal.Rptr. 515, 405 P.2d 555].) When the evidence "does not include an equally damaging confession," it is settled that the use of an inadmissible confession compels reversal. ( People v. Price, supra, 63 Cal.2d 370, 377.)
The judgment is reversed.
Assigned by the Chairman of the Judicial Council.
I concur in the reversal of the judgment, and in the determination that under state law and policy the holdings in Escobedo v. Illinois, 378 U.S. 478 [12 L.Ed.2d 977, 84 S.Ct. 1758], and People v. Dorado, 62 Cal.2d 338 [ 42 Cal.Rptr. 169, 398 P.2d 361], are partially retroactive and apply to cases that were not final on June 22, 1964, the date on which Escobedo was decided. I dissent, however, from the holding that the rules announced in Miranda v. Arizona, 384 U.S. 436 [16 L.Ed.2d 694, 86 S.Ct. 1602, 10 A.L.R.3d 974], under state law and policy, do not apply to pending appeals but only apply prospectively to cases tried after June 13, 1966, the date on which Miranda was decided. These two holdings are illogical, inconsistent, needlessly raise, without deciding, equal protection of the laws problems, and grant to some applicants while denying to others similarly situated basic constitutional rights.
In reaching the result that Miranda should only apply prospectively the majority follow Johnson v. New Jersey, 384 U.S. 719 [16 L.Ed.2d 882, 86 S.Ct. 1772], decided on June 20, 1966, in which the United States Supreme Court held that insofar as federal courts are concerned, Miranda only applied prospectively. In the same opinion the United States Supreme Court also held that, so far as federal law is concerned, the rules announced in Escobedo also had prospective operation only. But in Johnson v. New Jersey, supra, as the majority here points out, the United States Supreme Court, at some length, expressly pointed out that these federal rules were not binding on the states, and that the states were free to adopt other and different rules of retroactivity. That concession was undoubtedly specially noted because before Miranda a few states, including California, had held that Escobedo was retroactive to the extent of applying to pending appeals ( In re Lopez, 62 Cal.2d 368 [ 42 Cal.Rptr. 188, 398 P.2d 380]), and the United States Supreme Court wanted to assure those states that they could constitutionally adopt the same rule as to Miranda. This amounts to an open invitation to apply the same rules to Miranda as the states had applied to Escobedo. In my opinion we should accept that invitation. The majority repudiate it.
Whether we should or should not adopt the rules of retroactivity announced in Johnson v. New Jersey is now a matter of state policy — should we adopt the retroactivity rules of Johnson or the rules announced in Lopez? However we decide that question, it is submitted that the same rules should be adopted as to both Escobedo and Miranda, inasmuch as both deal with similar rights of freedom from self-incrimination and to counsel guaranteed by the federal and state Constitutions, and it is further submitted that the most compelling arguments are to the effect that we should adopt the rule of Lopez.
In determining whether or not to adopt, as a matter of state law and policy, the rules of retroactivity of Johnson v. New Jersey, we should examine the results and effects of that doctrine. It cannot be denied that the doctrine that rules announced in criminal cases involving fundamental constitutional rights shall operate prospectively is new and novel. I know of no precedent in our long judicial history where such a doctrine was even hinted at, far less decided. Certainly there is no California authority that supports any such doctrine. Such a doctrine, hinting as it does at legislative rather than judicial action, is completely foreign to our California law. Under our state law, prospective operation of judicial decisions has heretofore been limited to a very few civil cases and then only in most unusual cases. In most of the cases where the prospective operation rule has been adopted some element of estoppel was present. Thus to apply the doctrine to a criminal case involving constitutional rights is out of line with our state law. The United States Supreme Court acted well within its rule making powers when it made such a new and startling rule applicable to the federal courts. But, as already pointed out the states are under no compulsion to adopt such a doctrine, and some, like California, have been invited by the United States Supreme Court not to do so. Under such circumstances California should not blindly follow Johnson v. New Jersey as to the prospective operation of Miranda, and certainly should not do so after holding Escobedo applies to pending appeals. California should adopt a rule more in conformity with California law.
In considering what the state policy should be, we also have the right to consider the apparently conflicting and certainly anomalous holdings in Miranda and in Johnson v. New Jersey. In Miranda, and two of the cases decided along with it, the court reversed three convictions where the errors occurred before Miranda was decided. This necessarily constituted a determination that as to these three cases the rules announced in Miranda applied to pending appeals. This was on June 13, 1966. On June 20th the United States Supreme Court held the rules announced in Miranda should only apply after June 13, 1966, and dismissed many petitions for certiorari presenting the same problems as Miranda. Thus, if your name was Miranda, and your conviction had been secured by violation of rights set forth in Miranda, you were entitled to a reversal, but if your name was Johnson, or as in the instant case, Rollins, and your conviction was secured by methods condemned in Miranda, and you were tried at the same time or after Miranda was tried, but before June 13, 1966, your judgment must be affirmed. Although the instant case does not involve the death penalty, the same rules would be applicable there. Therefore a defendant in a murder case tried at the same time or after Miranda but before June 13, 1966, whose conviction rests on a confession secured in violation of Miranda, must die, but Miranda, under similar circumstances, would live. This presents, under the federal rule, an interesting and difficult equal protection of the laws problem. Although this is primarily the concern of the federal courts, we are entitled to consider, and should do so, those anomalies in determining whether we, as a matter of state policy, should adopt a rule that permits such inconsistencies.
The fourth case decided along with Miranda was People v. Stewart, a California case ( 62 Cal.2d 571 [ 43 Cal.Rptr. 201, 400 P.2d 97]). Among other things, the United States Supreme Court affirmed our decision where we had applied the Lopez rule.
There is another basic fundamental fallacy in the majority opinion. It adopts one rule of retroactivity as to Escobedo and another as to Miranda. The majority decide, and I agree, to adopt the rule of the Lopez case as a matter of state law, although when written we thought we were following federal law. But then the majority hold that Miranda applies only prospectively. The majority is right in adopting the rule of Lopez as to Escobedo. According to Dorado and Lopez, the rules announced in Escobedo were adopted in an attempt to preserve and protect Fifth and Sixth Amendment rights. The only criticism that can be made of the rule adopted in Lopez is that it did not go far enough — perhaps we should have held that the rule of retroactivity of Gideon v. Wainwright, 372 U.S. 335 [9 L.Ed.2d 799, 83 S.Ct. 792, 93 A.L.R.2d 733], was applicable. Certainly, strong arguments can be made for such a conclusion. But we decided, because of the impact on the administration of justice, to limit Escobedo to pending appeals, and to adopt a more conservative approach and gave to Escobedo only partial retroactivity. The view was later adopted by the United States Supreme Court in Linkletter v. Walker, 381 U.S. 618 [14 L.Ed.2d 601, 85 S.Ct. 1731], and Tehan v. Shott, 382 U.S. 406 [15 L.Ed.2d 453, 86 S.Ct. 459]. These cases tell us that constitutional rights, in certain cases, need only be applied to pending appeals. Unless we are compelled to do so we should not now hold that a violation of constitutional rights only adversely affects a defendant who is tried after June 13, 1966, and certainly, even if permitted to adopt such a rule, we should not do so when we have held correctly that Escobedo applies to pending appeals. Both deal with the same constitutional rights guaranteed by the Fifth and Sixth Amendments. Both deal with the same type of invasion of these rights — custodial interrogations with the warnings required. Consistency, logic, and the equal protection clause dictate that insofar as retroactivity is concerned Escobedo and Miranda be treated alike. Every argument of expediency made in the Johnson case and in the majority opinion applies to both legal situations. The argument of expediency made by the majority is totally without merit. This court in Lopez made Escobedo partially retroactive and the administration of justice managed to survive. The application of Escobedo to pending appeals resulted in many reversals and had a great impact on the police. Indisputably if Miranda were made applicable to pending appeals, reversals and hardships to the police would result. But these are false factors. The burdening of the administration of criminal justice is a price we must pay to enforce our Bill of Rights, and a price we should pay without hesitation. The powers of government, and that includes the police, are limited by the Bill of Rights. When such rights are invaded we must repulse such invasion and protect such rights even at the expense of inconvenience. Courts in the past have not hesitated to reverse convictions when constitutional rights have been invaded. We should do likewise.
The United States Supreme Court was at least being logical and consistent when it saw there were no basic differences in the rights involved in Escobedo and Miranda and held that the same rule as to retroactivity had to apply to both cases. The majority are illogical and inconsistent, and raise equal protection problems not discussed by them, when they hold one rule applies to Escobedo and another to Miranda. Logic and consistency may not always be attainable in the legal processes, but that does not mean that they should be disregarded, as they are in the majority opinion. When there is a logical and consistent solution to a problem, and an illogical and inconsistent one, the logical and consistent is to be preferred. We hold that Escobedo applies to pending appeals. Logically and consistently we should apply the same rule to Miranda. This, the majority does not do. As already indicated this raises a difficult question of equal protection that the majority do not discuss or even recognize.
The rights involved in Escobedo and Miranda — the right to counsel and the right not to incriminate oneself — are fundamental rights. The right to counsel has been declared to be essential to our system of ordered liberty. ( Palko v. Connecticut, 302 U.S. 319, 324-325 [82 L.Ed. 288, 58 S.Ct. 149].) Just recently the United States Supreme Court in Spevack v. Klein (Jan. 16, 1967) 385 U.S. 511 [17 L.Ed.2d 576, 87 S.Ct. 625], quoted with approval from Boyd v. United States, 116 U.S. 616, 635 [29 L.Ed. 746, 6 S.Ct. 524], where in discussing Fifth Amendment rights, the court stated, "It is the duty of courts to be watchful for the constitutional rights of the citizen, and against any stealthy encroachments thereon." Also in the Spevack case, supra, at page 516, the court in referring to the Fifth Amendment rights stated: "We find no room in the privilege against self-incrimination for classifications of people so as to deny it to some and extend it to others." This is precisely what the majority have done.
For these reasons while I concur in the reversal and in the holding that under state law and policy Escobedo should apply to pending appeals, I dissent from the holding that under state policy and law Miranda should only have prospective application. As a matter of state policy I think that Miranda, too, should have limited retroactivity.