¶1concurring.
¶2I concur in the result reached by the Court. I agree with the majority that “there is no presumed or per se rule of conflict of interest” regarding the representation of multiple defendants by staff attorneys at the same public defender’s office. Ante at 171.
¶3The majority’s opinion has additional significance. In resolving the claims of ineffective assistance of counsel in this case, the Court uses a balancing approach. It holds that a conflict of interest with resulting prejudice to defendants will be presumed only when “the circumstances demonstrate a potential conflict of interest and a significant likelihood of prejudice.” Id. 171. This holding confirms two salient points. The first is the soundness of a flexible rule for dealing with the prejudicial effects of professional conflicts of interest. The second is the converse of that proposition, namely, the imprudence and wastefulness of a per se rule that automatically imputes irrebuttable prejudice from any potential conflict of interest and then mandates reversal of a criminal conviction no matter how fairly founded and free of actual prejudice.
¶4Today’s decision is the third in a line of recent opinions by this Court on the subject of multiple representation. In State v. *176Land, 73 N.J. 24 (1977), a private attorney had represented codefendants at a criminal trial. In overturning the defendants’ convictions, we stressed the fundamental nature of the constitutional right to effective assistance of counsel. The Landcase presented a conflict of interest that was either actual or generated a substantial likelihood of prejudice. The same attorney represented a husband and wife, each of whom was charged separately with the commission of a joint crime. It is in that particular context that we held: “[I]n the absence of waiver, if a potential conflict of interest exists, prejudice will be presumed.” Id. at 35.
¶5The Landdecision was followed by State v. Bellucci, 81 N.J. 531 (1980), a case in which a private attorney and his partner had represented codefendants at. a joint trial. While presented with a less obvious conflict, this Court nevertheless gave the Landtest a sweeping application and ordered new trials for the defendants. Justice Pashman, writing for the majority, stated that “once a potential conflict exists, prejudice will be presumed in the absence of waiver.” 81 N.J. at 543.
¶6What emerged from the Belluccidecision was a per se rule against multiple representation that established an irrebuttable presumption of prejudice to the defendant and mandated the reversal of an otherwise sound criminal conviction simply because of the existence of a potential conflict of interest. I wrote separately in that case because I found the majority’s solution to the problems posed by multiple representation in criminal cases to be far too extreme. The Court’s methodology for dealing with multiple defendant representation required the judicial cancellation of otherwise fair criminal trials.
¶7In my estimation, these kinds of problems can be handled through a more tempered approach, which is entirely consonant with the constitutional interests that are implicated. In the absence of an actual conflict of interest, prejudice need be presumed only “where a potential conflict of interest is shown under circumstances generating a strong likelihood of actual *177prejudice to the defendant.” 81 N.J. at 547. This remedy accords a defendant sufficient constitutional protection without unnecessarily sacrificing the public’s strong interest in properly conducted, and concluded, criminal trials.
¶8The instant case illustrates the soundness and workability of such an approach. In this case staff attorneys in the same public defender’s office represented individual codefendants against criminal charges relating to the same criminal events or transactions. The entire Court—including the dissent—recognizes that the potential for conflicts of interest in this situation is comparable to the representation of multiple defendants by private attorneys who are members of the same law firm. I agree with this assessment. Each of these situations is fraught with professional conflicts which potentially jeopardize the criminal defendant’s constitutionally protected right to effective assistance of counsel. See Bellucci, 81 N.J. at 531; Land, 73 N.J. at 24; Cuyler v. Sullivan, 446 U.S. 335, 100 S.Ct. 1708, 64 L.Ed.2d 333 (1980).
¶9Nevertheless, in this case the Court refuses to presume actual prejudice and does not apply a per se rule of reversal. Ante at 167. The Court here uses essentially the same flexible approach that I urged in Bellucci. It recognizes the potential for prejudice inherent in the multiple representation of defendants by associated attorneys and then engages in a factual inquiry to determine whether actual or likely prejudice resulted from such multiple representation. Significantly, the Court ultimately determines that there was no actual or probable prejudice demonstrated in this case and, therefore, no reversible error. Ante at 171, 172-173.
¶10The Court also recognizes that while in this case it was confronted merely with a potential conflict, “should the circumstances demonstrate a potential conflict of interest and a significant likelihood of prejudice, the presumption of both an actual conflict of interest and actual prejudice will arise, without the necessity of proving such prejudice.” Ante at 171, 172-*178173. In my view, the Court’s general analysis and conclusions are eminently sound. My only difference with the majority concerns its continued adherence to a per se rule in regard to multiple representation by private attorneys. Ante at 171. Its reasons for doing so are not satisfactory.
¶11The Court departs from the automatic reversal rule of Bellucciin this case because the potential for conflict in a public defender’s office is more attenuated than in a private law firm. This reason is really nothing more than a pretext. I agree with Justice Pashman, who dissents on the ground that there is no material difference between public and private attorneys in terms of the potential conflict and prejudice. Ante at 167-169. Because these situations are sufficiently similar, Justice Pashman criticizes the majority for refusing to apply the Bellucci per se rule in this context. I criticize the majority for refusing to repudiate the Belluccirule.
¶12Despite its reaffirmance of Bellucci, the majority’s determination today can only be construed as a significant, albeit implicit, modification of its per se rule of automatic reversal. At the very least, however, the Court’s adoption of a different approach for public defenders, as opposed to private attorneys, exposes the weaknesses of its per se rule.
¶13An assumption underlying the Court’s decision is that there is a conflict of interest—to some degree—presented by the public defender’s representation of multiple defendants. The Court obviously considers the degree of potential conflict of interest to be more than de minimis. Otherwise, there would be no reason for its inquiry into whether there was “a significant likelihood of prejudice” caused by the conflict, however characterized. Ante at 171. ■
¶14Stripped to its essence, the Court’s opinion must be understood to hold that, whenever there is simply a potential for conflict, an inquiry into actual prejudice or the substantial likelihood of prejudice to a defendant’s right to effective assistance of counsel should be undertaken. While the presence of an actual *179conflict of interest or a conflict generating a substantial likelihood of actual prejudice, such as in Land, permits the imputation of prejudice, the Court’s opinion today demonstrates that where the conflict is not patent, an inquiry into actual or likely prejudice should be undertaken unless the asserted conflict is truly de minimis.
¶15Here, the factors that distinguish multiple representation by public defenders from multiple representation by private attorneys influence only the degree of the potentiality for conflicts of interest, not the existence vel non of such conflicts. The same concern for clients’ rights which we stressed in the Bellucciopinion presents itself with equal force whether the lawyers involved work at the same private law firm or the same public defender’s office. 81 N.J. at 541. See also Westinghouse Elec. Corp. v. Kerr-McGee Corp., 580 F.2d 1311, 1321 (7 Cir. 1978), cert. den., 439 U.S. 955, 99 S.Ct. 353, 58 L.Ed.2d 346 (1979). Logic, as well as sound policy, dictate that if all that is reasonably required in the one situation to protect a defendant’s rights is a factual inquiry into prejudice, that is all that need be undertaken in the other.
¶16In our eagerness to fashion strong rules for governing the handling of professional conflicts of interest in multiple defendant criminal trials, we can easily smother our objective. By formulating absolute rules that are so extreme and inflexible, we may be forfeiting other significant values in the administration of criminal justice. The polestar of the criminal justice system remains the fairness of the trial and the justness of the verdict. See State v. Singletary, 80 N.J. 55, 64 (1979); State v. Deatore, 70 N.J. 100, 105 (1976). I submit that a per se rule that requires the automatic reversal of a well-founded criminal conviction is self-defeating nullification of the criminal process. By clinging to its automatic reversal rule, the Court appears “to strain at a gnat and swallow a camel.” Holy Bible, Matthew 23:24 (King James Version).
¶17*180I do not gainsay the constitutional mischief that can arise in the unsupervised and uncorrected representation of multiple defendants. After a trial has been concluded, however, the focus must be on the nature, quality and degree of harmful prejudice suffered by a defendant as a result of multiple representation. It is only when a convicted defendant has suffered harmful injury to constitutional rights that he is entitled to another trial. See Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (167), reh. den., 386 U.S. 987, 875 S.Ct. 824, 17 L.Ed.2d 705; State v. Macon, 57 N.J. 325 (1971). The Supreme Court has sharpened this focus in Cuylerby positing as the standard for reversal of a conviction “an actual conflict of interest [that] adversely affected ... [the] lawyer’s performance.” 446 U.S. at 348, 100 S.Ct. at 1718, 64 L.Ed.2d at 346-47.
¶18Concededly, there are situations where the prejudice is not provable but, under the circumstances, is so patent or likely that a defendant should not be required to demonstrate the actual impairment of his constitutional rights. However, in the broad middle ground between a de minimis conflict and an actual conflict, where the existence of neither an actual conflict nor real prejudice is obvious, a more flexible standard can adequately protect the constitutional rights of the criminal defendant to effective assistance of counsel without requiring the needless reversal of otherwise valid criminal convictions. See, e.g., Wood v. Georgia, 450 U.S. 261, 270-72, 101 S.Ct. 1097, 1103-04, 67 L.Ed.2d 220, 230-31 (1981) (an apparent “possibility of conflict” imposes on the trial court a duty to inquire further); Cuyler, 446 U.S. at 350, 100 S.Ct. at 1719, 64 L.Ed.2d at 348 (Brennan, J., concurring) (a “significant possibility” of conflict creates a presumption of prejudice); Bellucci, 51 N.J. at 546, 547 (Handler, J., concurring) (a presumption of prejudice will arise where potential conflict of interest generates “a strong likelihood of actual prejudice”).
¶19It is hardly surprising, therefore, that this Court’s automatic reversal rule seems to stand alone, suggesting that it may indeed be aberrational. I am aware of no other jurisdiction that *181has opted for a per se rule requiring the automatic reversal of otherwise valid criminal convictions without a showing of actual or even probable prejudice. It does not appear that any other jurisdiction has adopted this absolutist approach to multiple representation situations when only a potential conflict of interest is presented. See, e.g., Cuyler, 446 U.S. 335, 100 S.Ct. 1708, 64 L.Ed.2d 333; United States v. Lawriw, 568 F.2d 98 (8 Cir. 1977), cert. den., 435 U.S. 969, 98 S.Ct. 1607, 56 L.Ed.2d 60 (1978); Haggard v. Alabama, 550 F.2d 1019 (5 Cir. 1977); Houston v. State, Ala.App., 401 So.2d 270 (1981); Sullivan v. State, Alaska, 509 P.2d 832 (1973); State v. Rogers, 110 Ariz. 582, 521 P.2d 1128 (1974); People v. Cook, 13 Cal.3d 663, 532 P.2d 148, 119 Cal.Rptr. 500, cert. den. 423 U.S. 870, 96 S.Ct. 135, 46 L.Ed.2d 100 (1975); People v. Romero, 189 Colo. 526, 543 P.2d 56 (1975); Palmer v. Adams, 162 Conn. 316, 294 A.2d 297 (1972); Isijola v. State, Del., 340 A.2d 844 (1975); Hall v. United States, D.C.App., 236 A.2d 57 (1967); State v. Youngblood, Fla., 217 So. 2d 98 (1968); Wright v. State, 158 Ga.App. 494, 280 S.E.2d 896 (1981); Pulver v. State, 93 Idaho 687, 471 P.2d 74 (1970); People v. Durley, 53 Ill.2d 156, 290 N.E.2d 244 (1972); Martin v. State, Ind., 314 N.E.2d 60, reh. den., Ind., 317 N.E.2d 430, cert. den., 420 U.S. 911, 95 S.Ct. 833, 42 L.Ed.2d 841 (1974); Jackson v. Aiger, Iowa, 239 N.W.2d 180 (1976); State v. Sullivan, 210 Kan. 842, 504 P.2d 190 (1972); Ware v. Commonwealth, Ky., 537 S.W.2d 174 (1976); State v. Johnson, La., 406 So.2d 569 (1981); Gayton v. Robbins, Me., 272 A.2d 776 (1971); Commonwealth v. White, 362 Mass. 193, 285 N.E.2d 110 (1972); State ex rel. Knott v. Tahash, 281 Minn. 305, 161 N.W.2d 617 (1968); Mooring v. State, Mo., 501 S.W.2d 7 (1973); State v. Jeffrey, 163 Mont. 92, 515 P.2d 364 (1973); Patterson v. State, 81 N.M. 210, 465 P.2d 93 (Ct.App.1970); People v. Crump, 53 N.Y.2d 824, 422 N.E.2d 815, 440 N.Y.S.2d 170 (1981); State v. Oliver, 23 Ohio App.2d 210, 262 N.E.2d 424, 52 Ohio Op.2d 308 (1970); Jones v. State, Okl.Cr., 527 P.2d 169 (1974); State v. Austin, 84 S.D. 405, 172 N.W.2d 284 (1969); Combs v. Turner, 25 Utah 2d 397, 483 P.2d 437 (1971); State v. Meyers, 86 Wash.2d 419, 545 P.2d 538 (1976); State ex *182rel. Postelwaite v. Bechtold, W.Va., 212 S.E.2d 69 (1975); State ex rel. White v. Gray, 57 Wis.2d 17, 203 N.W.2d 638 (1973).
¶20The Court’s conclusion in this case confirms the desirability of a flexible rather than an absolute rule. Such an approach in no way minimizes the importance of the underlying constitutional right. See Land, 73 N.J. at 24. Indeed, the Court’s opinion provides clarity and direction in resolving these matters. Ante at 174-175. There is a need for strong rules to deter conflicts of interest that threaten the right to effective assistance of counsel. These rules should, whenever possible, be invoked prior to trial and, if need be, before the conclusion of a trial. Properly followed and applied, such rules should suffice to assure defendants effective assistance of counsel, without the uncritical resort to the automatic imputation of prejudice and the concomitant reversal of criminal convictions.
¶21By following a balanced approach in this case, the Court has demonstrated that a flexible rule in multiple representation situations is feasible. Its decision illustrates that a judicial inquiry can properly be made to ascertain the nature of any asserted professional conflicts of interest and the extent of actual or probable prejudice to defendants’ rights engendered by that conflict. This approach better serves the fair, efficient and expeditious administration of criminal justice than does a per se rule that leads to automatic reversals of criminal convictions. I therefore urge the Court to extend this approach to multiple representation cases in general, rather than limit the holding to public defender cases.
¶22dissenting.
¶23The issue in this case is whether the representation of co-defendants by public defenders from the same office denies them effective assistance of counsel. In State v. Land, 73 N.J. 24 (1977), this Court held that representation of co-defendants by a single attorney was unconstitutional absent a valid waiver. Justice Schreiber, writing for a unanimous court, set forth our *183strong policy of avoiding even the possibility of conflicts of interest in the representation of co-defendants. In State v. Bellucci, 81 N.J. 531 (1980), the Court held that representation of co-defendants by two members of the same private law firm created a similar potential for conflict and therefore constituted a denial of effective assistance of counsel. Absent a knowing waiver, such joint representation is therefore prohibited. The majority here refuses to apply the same rule to public defenders from the same office. They believe that the salutary safeguards they have implemented to prevent actual conflicts, ante at 173-175, provide sufficient protection. In my view, extension of the Bellucei rule to public defenders is necessary to eliminate the danger of conflicts of interest. I therefore dissent.
¶24I agree with the majority that public interest attorneys and private attorneys should not automatically be treated the same way. Certain rules appropriate to one may be unnecessary or counterproductive when applied to the other. See, e.g., In re Education Law Center, Inc., 86 N.J. 124 (1981) (allowing nonprofit corporation with non-lawyer directors to practice law). The proper approach is to inquire into the appropriateness of applying the particular rule in question to public interest practice. Unlike the majority, I believe that anything short of applying the Bellucei rule to public defenders will not “protect adequately the ‘fundamental and absolute’ right to effective assistance of counsel.” State v. Bellucei, 81 N.J. at 543 (citation omitted).
¶25The majority’s approach centers around a hearing early in the proceedings to determine whether there is any potential for conflict. See R. 3:8-2. If so, joint representation will be prohibited. If defense counsel perceives a potential conflict, the trial court is urged, though not required, to defer to that judgment. There are two problems with that solution. First, it leaves open the danger that a conflict will be discovered later in the proceedings, perhaps during the trial. Second, it does nothing to address the fear some defendants may have of revealing confidences to an attorney associated with the attorney for their *184co-defendants. The Belluccirule would fully eliminate both problems.
¶26Conflicts of interest will not always be evident at the time of the pre-trial hearing. As Justice Schreiber pointed out in State v. Land, 73 N.J. at 32, “[although not apparent, the conflict may surface during the course of the trial.” The facts of this case underscore that possibility. At the last minute, the State discovered and introduced a witness whose testimony was damaging to only one of the co-defendants. Although I do not dispute the majority’s finding that there was no actual prejudice here, it is easy to see how on slightly different facts the introduction of such a witness could create a conflict.
¶27If a conflict is discovered during or immediately before trial, the defense and the court are faced with two unacceptable alternatives. The first possibility is that they may choose to tolerate the conflict rather than change attorneys. Defense counsel who has put considerable work into a case may hesitate to see that work go to waste and burden an equally busy neighboring public defender office with a duplication of effort. A defendant who has confided in and learned to trust his attorney may view the conflict of interest as the lesser evil when faced with the prospect of a new attorney. Finally, especially during trial, the trial court may view a defense motion for a change in attorneys as an attempt to stall or disrupt the trial. Its inclination may therefore be to deny the motion. In each case, the result will be ineffective assistance of counsel.
¶28The alternative is to change attorneys in midstream. This is equally unacceptable. At best, such a change would require substantial time and resources for the new attorney to familiarize herself with the case. The trial would have to be delayed during this time or, if the trial had already begun, a new trial might be necessary. Often, the result may be even worse. A new attorney may be appointed and required to proceed immediately with the case, without the opportunity to do much investí*185gation, study the files of the prior attorney, or even talk with her client.
¶29This solution will also have an adverse impact on the defendant’s perception of the criminal justice system. He will be forced to change attorneys in the middle of his case. Any confidence he has gained in his attorney, any feeling that he can trust and confide in her, will be dissipated. He will be forced to reveal difficult and personal facts about himself to still another stranger. People close to the criminal justice system are well aware that one prevalent complaint of criminal defendants is that they have been shunted from attorney to attorney during the course of their case. Without an opportunity to develop any sort of rapport with their attorneys, defendants often feel that they have not had an adequate opportunity to present a defense. Often, they will be right. Such shuffling is certainly inconsistent with the type of individualized attention and zealous advocacy that our criminal justice system purports to provide defendants. That the majority’s rule will unavoidably increase the shuffling of appointed defense attorneys is alone sufficient reason to reject it.
¶30The second difficulty with the majority’s rule concerns client confidences. I assume that hardly ever would a public defender reveal her client’s confidences to another public defender in the same office representing a co-defendant. Even so, the defendant’s perception that his attorney will share information with an associate can damage the attorney-client relationship almost as much as the actual revelation of such information. We have said that “the right to counsel ‘would be meaningless if the defendant were not able to communicate freely and fully with the attorney.’ ” State v. Land, 73 N.J. at 30, quoting M. Freedman, Lawyers’ Ethics In An Adversary System 8 (1975). Such open communication will not occur if the defendant believes that his secrets will not be safe, whether or not that belief is justified.
¶31*186The facts of State v. Rogers, 90 N.J. 187 (1982), decided today, illustrate the problem. The defendant there believed that anything he said to his attorney would be disclosed to the attorney for his co-defendant. He therefore refused to discuss his case with his attorney. The trial court determined that his fears about confidentiality were unfounded and therefore denied his motion for a new trial. However, the question is not whether his fears were justified. Whether or not they were justified, they were real. Given those fears, defendant could not communicate with his attorney. Since he was not given an attorney he felt he could trust, he was denied effective assistance of counsel.
¶32The right to independent counsel is fundamental. As we explained in Bellucci, supra:
The harm in dual representation is caused by the restraints placed on an attorney’s advocacy and independent judgment. It is one of divided loyalties. At its extreme, such conflict may prevent counsel from attempting to exonerate one client when doing so would require him to demonstrate that another client is guilty. [81 N.J. at 543]
¶33The majority offers no reason why co-defendants need be represented by public defenders from the same office despite the potential for conflict. Any savings in resources from the majority’s rule likely will be outweighed by the added expense and duplication of effort that will result from changing attorneys in midstream as conflicts arise. In any event, convenience and reduction of administrative costs cannot justify deprivation of the effective assistance of counsel.
¶34*187My disagreement with the majority is not at the level of principle. We all agree that the constitutional right to effective assistance of counsel deserves careful protection. The majority believes that their rule will adequately protect that right. I disagree. In my view, application of the Belluccirule to public defenders from the same office is necessary to assure defendants their fundamental right to independent counsel.
¶35HANDLER, J., concurring in the result.
¶36For affirmance —Chief Justice WILENTZ and Justices CLIFFORD, SCHREIBER, HANDLER, POLLOCK and O’HERN—6.
¶37Dissenting—Justice PASHMAN—1.
¶38I understand that the financial costs of extending the Belluccirule are not negligible. For that reason, whenever possible, co-defendants should be represented by public defenders from neighboring municipalities rather than by pool counsel. In this way, there will be no additional fees.