(dissenting) — I cannot agree it is "harmless” error to allow the jury to consider evidence otherwise *329inadmissible because it is unduly prejudicial.3 Prejudicial evidence is not harmless evidence.
The "harmless error” doctrine requires reversal whenever there is a reasonable likelihood that inadmissible evidence affected the jury’s verdict. State v. Owens, 128 Wn.2d 908, 914, 913 P.2d 366 (1996); State v. Copeland, 130 Wn.2d 244, 287, 922 P.2d 1304 (1996). Conclusions of harmless error based upon this reasonable likelihood inquiry amount to little more than speculative appellate fact-finding.
Jurors and courts are made up of human beings, whose condition of mind cannot be ascertained by other human beings. Therefore, it is impossible for courts to contemplate the probabilities any evidence may have upon the minds of the jurors. The state attempts to safeguard the life and liberty of its citizens by securing to them certain legal rights. These rights should be impartially preserved. They cannot be impartially preserved if the appellate courts make of .themselves a second jury and then pass upon the facts.
State v. Robinson, 24 Wn.2d 909, 917, 167 P.2d 986 (1946) (emphasis added). See also State v. Finnegan, 6 Wn. App. 612, 622, 495 P.2d 674 (quoting State v. Macon, 57 N.J. 325, 273 A.2d 1, 9 (1971)), review denied, 81 Wn.2d 1001 (1972), cert. denied, 410 U.S. 967, 93 S. Ct. 1450, 35 L. Ed. 2d 702 (1973).
*330Both state and federal constitutions guarantee the right to have a jury resolve factual issues. Coppo v. Van Wieringen, 36 Wn.2d 120, 121, 217 P.2d 294 (1950) (citing Wash. Const, art I, § 21); U.S. Const, amend. VI. Yet the majority uses the harmless error doctrine to displace the jury with unreflecting alacrity. State v. Pam, 30 Wn. App. 471, 477, 635 P.2d 766 (1980) (Ringold, J., dissenting), aff’d, 98 Wn.2d 748, 659 P.2d 454 (1983). Such judicial "second-guessing” rests in pure conjecture as this court can never be privy to precisely what evidence appealed to individual jurors much less the jury as a collective body. See State v. Valentine, 132 Wn.2d 1, 21, 935 P.2d 1294 (1997); Phillips v. City of Seattle, 111 Wn.2d 903, 911-12, 766 P.2d 1099 (1989); Dennis J. Sweeney, An Analysis of Harmless Error In Washington: A Principled Process, 31 Gonz. L. Rev. 277 (1995-96).
Even assuming a "reasonable likelihood” inquiry is ever appropriate, this error was certainly not harmless. This case warrants reversal because here there is much reason to believe this admittedly prejudicial evidence likely affected this verdict.
The court must presume the jury considered all the evidence before it, Zukowsky v. Brown, 79 Wn.2d 586, 603, 488 P.2d 269 (1971), and courts generally recognize psychiatric testimony impresses a jury. Ake v. Oklahoma, 470 U.S. 68, 81-82, 105 S. Ct. 1087, 84 L. Ed. 2d 53 (1985) (quoting F. Tee Bailey & Henry B. Rothblatt, Investigation and Preparation of Criminal Cases § 175 (1970)). Credible psychiatric testimony is critical to prove the defense of diminished capacity. Id. When inadmissible evidence clearly undermines the credibility of an expert witness and occupies the center of a party’s case, much more than a reasonable probability exists that such evidence affected the verdict. How can the majority confidently claim otherwise?
Here the majority finds it was inherently prejudicial to *331inform the jury that the source of adverse prosecution testimony was an expert originally retained by the accused, Majority at 326; yet the majority concludes it was "harmless” error to do so. How something can be inherently prejudicial and harmless all at the same time is beyond my comprehension.
Three psychiatric experts testified at trial. Of those, the disclosure affected the credibility of at least two by enhancing the credibility of Dr. Harris for the prosecution at the expense of the credibility of Dr. Liebert for the defense. The majority admits such disclosure clearly invited the jury to infer Dr. Harris was the more credible and encouraged the jury to give additional weight to his testimony at the expense of the defense witness. Majority at 327. Disclosure also suggested the defense hid relevant information. In effect it actually forced the defense to "involuntarily vouch” for Dr. Harris’ credibility. See United States v. Walker, 910 F. Supp 861, 864 (N.D.N.Y. 1995).
Informing the jury Dr. Harris was first retained but then rejected as a defense witness likely impugned the testimony of Dr. Liebert as well. By inferring the defense had to "shop” for a favorable opinion the jury could have easily inferred Dr. Liebert testified simply as instructed by the defense, thus further undermining his credibility.
Claiming—actually speculating—the error was harmless, the majority argues the accused compromised his credibility by allegedly failing to tell Dr. Harris about his flashbacks. Majority at 327-28. The majority also asserts the defendant was angry or jealous of Mr. Washington and that Dr. McFall’s testimony precludes any reasonable probability that the jury’s verdict was affected. Majority at 328. But the jury might conclude otherwise. The defendant testified he did not disclose his flashbacks to Dr. Harris out of embarrassment. More importantly, the defendant testified he experienced traumatic events during his *332service in Vietnam and as a police officer.4 He was diagnosed with posttraumatic stress disorder, a disorder which could have caused dissociation prior to the assault. Finally, the testimony of Dr. McFall, the third expert at trial, does not eliminate the prejudicial impact of the inadmissible evidence as the inadmissible evidence affected the testimony of two out of three experts.
In reality we cannot second-guess the jury nor can we determine precisely what evidence affected the verdict in what way. But we can be certain that psychiatric expert testimony played a crucial role in the outcome, and the inadmissible disclosure went to the very heart of the defense. For these reasons, I cannot conclude the error was merely harmless and would reverse and remand this case for a new trial.
I concur in the majority’s decision that State v. Pawlyk, 115 Wn.2d 457, 800 P.2d 338 (1990), governs this case and permits disclosure of the name of the non-testifying defense expert, his written reports, tests, notes, and oral interview. Pawlyk also allows the State to call that expert as a rebuttal witness. However the wisdom of Pawlyk is questionable.
Stare decisis requires "[o]nce this court has decided an issue of state law, that interpretation is binding until we overrule it.” Hamilton v. Department of Labor & Indus., 111 Wn.2d 569, 571, 761 P.2d 618 (1988). Therefore for Pawlyk to be overruled as the Washington Association of Criminal Defense Lawyers requests, a clear showing must be made that the "established rule is incorrect and harmful before it is abandoned.” In re Stranger Creek, 77 Wn.2d 649, 653, 466 P.2d 508 (1970). Pawlyk’s original dissent persuasively argued the criminal rules do not support this type of discovery and the attorney-client privilege and work-product doctrines prevent such disclosure. Moreover, in retrospect, the factual predicate of Pawlyk insofar as it relates to multiple psychiatric evaluations is based upon pure conjecture and unsupported speculation. See Pawlyk, 115 Wn.2d at 462-63.1 doubt it would pass scrutiny under Frye v. United States, 293 F. 1013, 34 A.L.R. 145 (D.C. Cir. 1923).
Defendant served as a United States Marine Corps rifleman in Vietnam. During his tour of duty, he was wounded three times in combat and received three Purple Hearts. He also served as a Seattle police officer for over 19 years.