dissenting.
The majority holds that under federal law, as articulated by the United States Supreme Court in Arizona v. Fulminante, _ U.S. _, 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991), harmless error analysis is applicable to every case wherein a defendant demonstrates a denial of the federal constitutional right to be present at trial.1 Op. at 275. I disagree with this reading of Ful-minante and conclude that harmless error is not applicable in the context of this case. I therefore respectfully dissent from the majority’s opinion.
In Fulminante, the Supreme Court affirmed a judgment of the Arizona Supreme Court holding that the admission into evidence of a coerced confession required a new trial. A bare majority of the United States Supreme Court (hereafter referred to as “the Fulminante majority”) held that harmless error analysis was applicable to cases involving the admission into evidence of coerced confessions.2 Fulminante, _ *281U.S. at _ - _, 111 S.Ct. at 1263-66. Four members of the Court concluded that harmless error analysis was never applicable to coerced confession cases. Id. at _ - _, 111 S.Ct. at 1253-57. Given the division of the Court over this particular issue, it is especially important to review the reasons articulated by the Fulmi-nante majority for the conclusion that harmless error analysis is available in this class of cases.
The Fulminante majority initially observes that the Court has applied harmless error analysis “to a wide range of errors and has recognized that most constitutional errors can be harmless.” Id. at _, 111 S.Ct. at 1263. The opinion then lists sixteen prior United States Supreme Court decisions as illustrative of the foregoing observation. Six of the cases dealt with improper jury instructions. Seven of the cases concerned issues relating to the exclusion or admission of evidence. One case dealt with improper prosecutorial comment on the defendant’s failure to testify, and one case involved a denial of counsel at a preliminary hearing. The remaining case, Rushen v. Spain, 464 U.S. 114, 104 S.Ct. 453, 78 L.Ed.2d 267 (1983), is viewed by the majority in the instant case as dispositive of the issue before us.
The Fulminante majority then enunciates a standard by which to judge the availability of harmless error analysis in criminal cases wherein a defendant has been denied a constitutionally protected right. The standard distinguishes between “trial errors” and “structural defects.” The former term refers to errors “which occurred during the presentation of the case to the jury, and which may therefore be quantitatively assessed in the context of other evidence presented in order to determine whether its admission was harmless beyond a reasonable doubt.” Fulminante, _ U.S. at _, 111 S.Ct. at 1264. Structural defects are “defects in the constitution of the trial mechanism” and are considered errors “affecting the framework within which the trial proceeds, rather than simply an error in the trial process itself.” Id. at _, 111 S.Ct. at 1265. The Fulminante majority then concludes that the admission of a coerced confession constitutes a trial error and is therefore subject to harmless error analysis.
Application of this dichotomy to the issue presented here requires the conclusion that in this case the deprivation of the defendant’s right to be present at this concededly critical stage of the trial is not a trial error susceptible of quantitative analysis, but rather constitutes a structural defect in the trial mechanism which by its very nature precludes harmless error analysis. Unlike cases involving evidentiary errors or errors in instructions, the absence of a defendant’s presence at a critical stage of trial proceedings is rarely subject to any quantitative assessment. In the former class of cases, the record may be examined to assess the impact of the constitutional error upon the jury in light of other evidence or other instructions also considered by the jury. In such a case the record of the proceedings permits an objective determination of whether the proceedings were fair beyond a reasonable doubt. The record here, however, contains no information permitting objective assessment of the impact of the defendant’s absence on the conduct of his attorney or on the jury.
As a result of the defendant’s absence, his attorney had no opportunity to consider questions, concerns, observations or suggestions offered by the defendant in reaction to statements made by the prosecution during the prosecutor’s initial closing argument. While the record contains the prosecutor’s initial closing argument, nothing in the record permits evaluation of the defendant’s reactions to that argument or the significance of any observations the defendant may have imparted to his attorney about the prosecutor’s statements. In addition, the psychological effect on the jury of the defendant’s inability to comprehend or comment upon this critical stage of the trial is not capable of quantitative analysis. See Larson v. Tansy, 911 F.2d 392, 395-96 (10th Cir.1990). Under these circumstances, it is not possible to measure the impact of this structural defect. See Hays v. Arave, 977 F.2d 475 (9th Cir.1992) (denial of defendant’s right to be present at sen*282tencing proceeding is structural error under Fulminante and is not subject to harmless error analysis).
The majority here relies upon the United States Supreme Court decision in Rushen v. Spain, 464 U.S. 114, 104 S.Ct. 453, 78 L.Ed.2d 267 (1983), in support of the conclusion that “allegations of denial of the right to be present at trial are scrutinized under the harmless error doctrine.” Opinion at 274. Rushen was among those cases cited by the Fulminante majority as illustrative of trial errors. Fulminante, _ U.S. at _, 111 S.Ct. at 1263. In Rushen, a juror approached the trial judge on two occasions during the course of a lengthy trial with questions regarding the appropriateness of the juror’s continued service. The trial judge briefly responded, but did not disclose these ex parte communications to the defendant's attorney. Upon discovering these facts after the jury had rendered its verdict, the defendant’s attorney unsuccessfully moved for a new trial. On appeal, the California Court of Appeals affirmed the conviction, concluding that the ex parte communications constituted harmless error.
The defendant sought habeas corpus relief from the federal courts. A federal district court ordered the defendant’s release, concluding that the absence of a contemporaneous record made it impossible to apply harmless error analysis. Spain v. Rushen, 543 F.Supp. 757, 770 (N.D.Cal.1982). On appeal, the Ninth Circuit Court of Appeals affirmed, holding that such an unrecorded ex parte communication could never constitute harmless error. Spain v. Rushen, 701 F.2d 186 (9th Cir.1983) (mem.).
In Rushen, six members of the Supreme Court, in a per curiam opinion, reversed the judgment of the Ninth Circuit Court of Appeals. The per curiam opinion “emphatically” disagreed with the Ninth Circuit’s conclusion that an unrecorded ex parte communication between a trial judge and a juror can never be subject to harmless error analysis. Rushen, 464 U.S. at 117, 104 S.Ct. at 455. In a footnote the Court, having observed that the case before it implicated both right to counsel and right of presence issues, stated that “[tjhese rights, as with most constitutional rights, are subject to harmless-error analysis, unless the deprivation, by its very nature, cannot be harmless.” Id. at 117-18 n. 2, 104 S.Ct. at 455 n. 2 (citations omitted). This statement recognizes that some deprivations of the right to be represented by counsel and the right to be present during critical stages of trial proceedings are by their very nature not amenable to harmless error analysis, in view of this language, I cannot agree with the majority’s view that Rushen stands for the proposition that in all circumstances a deprivation of the right to be present at a critical stage of trial proceedings must be scrutinized under the harmless error doctrine.
In this ease, the defendant was denied his fundamental right to be present at a critical stage of the criminal proceedings brought against him. The record does not contain any information that would permit a quantitative assessment of the impact of such deprivation on the fairness of the deliberative process. In my view, Fulmi-nante requires the conclusion that in these circumstances the error must be considered a structural defect and is not susceptible to harmless error analysis. I therefore would reverse the judgment of the Court of Appeals and direct that court to remand the case to the trial court for a new trial.
. In this case it is conceded that the defendant was denied his constitutionally protected right to be present at all critical stages of trial. We are thus not concerned with questions of waiver or whether the giving of instructions, the delivery of closing arguments and the receipt of the jury’s verdict should be categorized as critical stages of the trial process.
. A different majority of the Court concluded that admission of the coerced confession was not harmless beyond a reasonable doubt. Of course, the case before us does not involve a coerced confession.