Judge ERDMANN
delivered the opinion of the court.
Sergeant (SGT) William J. Kreutzer Jr. opened fire with an automatic weapon on personnel in his brigade when they were in formation commencing a unit run. He was subsequently charged with one specification of premeditated murder, eighteen specifications of attempted premeditated murder, one [295]*295specification of violation of a lawful general regulation, one specification of larceny of Government munitions, four specifications of maiming, and eighteen specifications of aggravated assault, in violation of Articles 118, 80, 92, 121, 124, and 128, Uniform Code of Military Justice (UCMJ), 10 U.S.C. §§ 918, 880, 892, 921, 924, 928 (2000), respectively. The charges were referred to a general court-martial with instructions that the case was “[t]o be tried as a capital case.”
Kreutzer pleaded guilty to one specification of murder while engaged in an act inherently dangerous to another (as a lesser included offense of premeditated murder), eighteen specifications of assault with a loaded firearm (as a lesser included offense of attempted premeditated murder), one specification of violating a lawful general regulation, and one specification of larceny of Government munitions. He was convicted of one specification of premeditated murder, eighteen specifications of attempted premeditated murder, one specification of violating a lawful general regulation, and one specification of larceny of Government munitions.1 A unanimous twelve-member court of officer and enlisted members sentenced Kreutzer to death, a dishonorable discharge, forfeiture of all pay and allowances, and reduction to E-l. The convening authority approved the sentence as adjudged.
The United States Army Court of Criminal Appeals affirmed only the findings to which Kreutzer had entered guilty pleas: murder while engaged in an inherently dangerous act; assault with a loaded firearm; violating a lawful general regulation; and larceny of Government munitions. United States v. Kreutzer, 59 M.J. 773, 784 (A.Ct.Crim.App.2004). The Army court set aside the findings of guilty to premeditated murder and attempted premeditated murder and the sentence. Id. The Army court denied a Government request for en banc consideration and a motion for reconsideration. Subsequently, the Judge Advocate General of the Army (TJAG) certified the case to this court pursuant to Article 67(a)(2), UCMJ, 10 U.S.C. § 867(a)(2) (2000).
Compulsory process, equal access to evidence and witnesses, and the right to necessary expert assistance in presenting a defense are guaranteed to military accuseds through the Sixth Amendment, Article 46, UCMJ, 10 U.S.C. § 846 (2000), and Rule for Courts-Martial (R.C.M.) 703(d). See Ake v. Oklahoma, 470 U.S. 68, 77, 105 S.Ct. 1087, 84 L. Ed.2d 53 (1985). The Court of Criminal Appeals found that Kreutzer was erroneously denied expert assistance in the form of a capital mitigation specialist. Kreutzer, 59 M. J. at 775. A majority of that court further found that the Government did not show that error to be harmless beyond a reasonable doubt with respect to the contested findings of premeditated murder and attempted premeditated murder. Id. The issue certified to us by TJAG asks us to determine whether the Court of Criminal Appeals erred in finding that the Government did not meet its burden of demonstrating that the erroneous denial of a mitigation specialist was harmless beyond a reasonable doubt.2
[296]*296BACKGROUND
Sergeant Kreutzer enlisted in the U.S. Army and entered active duty in February 1992. In March of 1993 he was assigned to the 325th Airborne Infantry Regiment of the 82d Airborne Division,3 stationed at Fort Bragg, North Carolina. Although Kreutzer was considered by some superiors to be a good soldier, throughout his military career he had trouble fitting in and interacting with his fellow troops. Kreutzer deployed to the Sinai with his unit in January 1994. The butt of numerous practical jokes and frequently referred to in derogatory terms, Kreutzer’s tolerance for this chiding and his relations within his unit deteriorated. From approximately April to July, 1994, Kreutzer began to make threats to kill including threats to kill named individuals, soldiers in physical training formation, and entire units while they slept. Kreutzer often described to others precisely how he planned to kill these individuals.
In June of 1994 Kreutzer broke down and cried while on guard duty and threatened to kill other members of his unit. Kreutzer was confronted about the threats by his platoon sergeant who escorted him to the division’s mental health officer, Dr. (Captain) Darren Fong. Kreutzer met with Dr. Fong and discussed his homicidal thoughts about his squad members. Doctor Fong concluded that Kreutzer was a person who had problems with anger and low self-esteem and “appeared” to have interpersonal problems with members of his squad. Doctor Fong, however, felt that Kreutzer was not a danger to himself or others. Kreutzer remained with his unit where he continued to have difficulty interacting with other soldiers and also started to have performance problems that continued up to the time of the charged offenses. Despite his difficulties and apparent emotional problems, in October 1994, Kreutzer attended the Primary Leadership Development Course to become a noncom-missioned officer and he was promoted to sergeant in March 1995.
The 2d Brigade scheduled a unit run for the morning of October 27, 1995.4 On the evening of October 26, 1995, Kreutzer called Specialist Fourth Class (SP4) Mays, a member of his squad, and informed him that he was “going to shoot the run the following day.” Before the scheduled run on the morning of October 27, SP4 Mays informed his chain of command of Kreutzer’s threat. The platoon leader and platoon sergeant initially laughed when the threat was brought to their attention. Similarly, the threats Kreutzer made to SP4 Mays were not taken seriously by the first sergeant or the acting company commander.
On the morning of October 27, 1995, Kreutzer secreted himself in the woods near an athletic field at Fort Bragg, North Carolina, where his brigade was forming for the run. As the brigade marched out of the stadium, Kreutzer opened fire on the formation with two rifles. Eighteen soldiers were wounded and one officer suffered a fatal wound in Kreutzer’s attack. Kreutzer was subdued by three Special Forces soldiers who had been running in the area.
Kreutzer did not deny that he fired shots at his brigade from ambush or that he wounded eighteen soldiers and killed an officer. He later stated that his actions were intended to send a message that the unit did not care about the men. He also anticipated [297]*297that he would either shoot himself or be shot and killed by the military police.
Subsequent to his apprehension and while still in law enforcement custody, Kreutzer was assessed by the 82d Airborne Division psychiatrist. He received additional mental health evaluation in the form of a suicide assessment while he was in pretrial confinement. Kreutzer was evaluated by a privately retained forensic psychiatrist in November 1995 and he was evaluated by a sanity board convened under R.C.M. 706 in December 1995. Charges against Kreutzer were referred to a capital general court-martial on January 24,1996.
On March 11, 1996, Kreutzer’s defense team filed a request with the convening authority for employment of a mitigation specialist. Although the convening authority did approve funding for defense counsel to travel in support of building a case in mitigation, he denied the request for a mitigation specialist. The defense renewed its request for a mitigation specialist by motion before the military judge.
Defense counsel provided a copy of the request they had made to the convening authority in which they asserted that they lacked “the experience and scientific expertise to uncover all potentially mitigating events or factors in SGT Kreutzer’s case.” They also provided an extensive affidavit from a mitigation specialist that explained the necessity of a mitigation investigation in capital cases, the scope of that investigation, and the role of a mitigation specialist. Defense counsel argued that they were not qualified to do the work of a mitigation expert. The Court of Criminal Appeals reached the same conclusion stating, “[I]n determining whether or not the defense counsel could adequately do the function of an expert mitigation specialist, the judge should have considered, among other factors, defense counsel’s lack of capital litigation experience, their minimal capital litigation training, and the time constraints they were facing at trial.” Kreutzer, 59 M.J. at 778. The military judge denied the motion without entering any findings of fact by simply stating: “I find the law here at Loving 3 at 250. I don’t find the showing requiring me to order one.”5
DISCUSSION
I.
Ruling of the U.S. Army Court of Criminal Appeals
Kreutzer appealed a number of issues to the Court of Criminal Appeals,6 but the Army court addressed only two: “(1) whether the military judge erred by denying [Kreutzer] the services of an expert consultant in capital sentence mitigation, and (2) whether appellant’s detailed trial defense counsel were ineffective in their representation of [Kreutzer] at the sentencing stage of trial.” Id. at 775. The court unanimously agreed that the sentence must be set aside due to ineffective assistance of counsel, a determination that the Government has not appealed. A majority of the court found prejudicial error in the denial of the mitigation specialist and set aside the contested findings and sentence. The Government does not contest that the military judge erred in denying a mitigation specialist, but argues that the error was harmless beyond reasonable doubt for the following reasons: (a) the Court of Criminal Appeals erred in applying the test for harmless error; (b) the majority opinion impeached itself; and (c) the nature of the mental health evidence is not relevant to the element of premeditation. Following discussion of the Government’s contentions in regard to the Army court’s decision, we will address whether the Government met its burden in establishing that Kreutzer suffered no prejudice.
[298]*298a. Did the Court of Criminal Appeals err in applying the test for harmless error?
Arguments:
The Government argues that the Court of Criminal Appeals erred in applying the test for harmless error in this case. The Army court explained that the test for prejudice was whether the error was harmless beyond a reasonable doubt. Kreutzer, 59 M.J. at 779 (citing Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967)), and went on to define that inquiry as whether the error itself had substantial influence on the trial results. Id. (citing United States v. Pollard, 38 M.J. 41, 52 (C.M.A.1993)). The Government urges that Pollard requires a threshold determination that an error must be shown to “prejudice a litigant’s substantial rights” before any burden to show harmlessness beyond a reasonable doubt is imposed on the Government. 38 M.J. at 52 (quoting Kotteakos v. United States, 328 U.S. 750, 760, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946)). The Government argues that Kreutzer did not meet this threshold showing of prejudice.
Kreutzer argues that the Court of Criminal Appeals used the correct standard in determining this error was not harmless. In response to the Government’s Pollard argument, Kreutzer argues that the Army court did find prejudice to his substantial rights and, alternatively, that the finding that the error was not harmless beyond a reasonable doubt created a de facto finding that the error prejudiced his substantial rights. Finally Kreutzer points out that this is a case in which the test for harmlessness must be assessed by asking if even one member might have been influenced to conclude that the Government did not prove premeditation beyond a reasonable doubt. Had even one member been so persuaded, death would have been removed as a lawful punishment in this case.
Discussion:
The right to the expert assistance of a mitigation specialist in a capital case is determined on a case-by-case basis. See United States v. Loving, 41 M.J. 213, 250 (C.A.A.F.1994).7 Where such a request is erroneously denied, that ruling implicates the right to present a defense, compulsory process, and due process conferred by the Constitution, the right to obtain witnesses and evidence conferred by Article 46, UCMJ, and the right to the assistance of necessary experts conferred by R.C.M. 703(d). Because Kreutzer was wrongly deprived of the expert assistance of a mitigation specialist to aid in the preparation of this capital case, the Court of Criminal Appeals held that Kreutzer had been denied due process. Kreutzer, 59 M.J. at 779. Concerning findings, the lead opinion of the Court of Criminal Appeals concluded: “Here the judge’s abuse of discretion adversely impacted the fairness of the trial on findings as to the issue of premeditation by depriving appellant of a complete presentation of the evidence concerning his state of mind____” Id. at 779-80. The Government does not contest the finding that Kreutzer was denied due process. This error of constitutional magnitude must be tested for prejudice under the standard of harmless beyond a reasonable doubt. See Chapman, 386 U.S. at 24, 87 S.Ct. 824; United States v. Sidwell, 51 M.J. 262, 265 (C.A.A.F.1999). The inquiry for determining whether constitutional error is harmless beyond a reasonable doubt is “whether, beyond a reasonable doubt, the error did not contribute to the defendant’s conviction or sentence.” United States v. Kaiser, 58 M.J. 146, 149 (C.A.A.F.2003) (quoting United States v. Davis, 26 M.J. 445, 449 n. 4 (C.M.A.1988)). We will reverse a conviction unless we find that a [299]*299constitutional error was not a factor in obtaining that conviction.
The Court of Criminal Appeals found that the denial of a mitigation specialist was a denial of due process and articulated the requirement that a conviction cannot be affirmed if a constitutional error was not harmless beyond a reasonable doubt. Kreutzer, 59 M.J. at 779-80 (citing Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967)). This is “a familiar standard to all courts.” Chapman, 386 U.S. at 24, 87 S.Ct. 824. Nonetheless, the Court of Criminal Appeals went on to misstate the nature of the inquiry.
The Army court defined the constitutional error inquiry as follows: “In testing for harmless error we inquire ‘whether the error itself had substantial influence’ on the trial results.” Kreutzer, 59 M.J. at 779 (quoting Pollard, 38 M.J. at 52; Kotteakos, 328 U.S. at 765, 66 S.Ct. 1239). In Pollard we assessed the impact of a trial counsel erroneously reading a victim’s pretrial statement to the members in the guise of impeachment — a nonconstitutional trial error. 38 M.J. at 50-51. That assessment was to determine if “an error of law ... materially prejudiced] the substantial rights of the accused.” Pollard, 38 M.J. at 51-52; Article 59(a), UCMJ, 10 U.S.C. § 859(a) (2000). In contrast to asking whether a constitutional error contributed to a conviction, the quest for a “substantial” influence seeks a more measurable impact or importance. When constitutional error is substantial and, as reflected by Chapman, where that error contributes to a conviction, the conviction cannot stand. We hold that in relying upon Pollard and testing this error for “substantial influence,” the Army court applied an erroneous definition to the nature of the inquiry into the effect of constitutional error.8 The error is not material to this appeal, however, because the standard that the Army court applied found harm under a test more favorable to the Government than the standard it should have applied.
This court reviews de novo the issue of whether a constitutional error was harmless beyond a reasonable doubt. United States v. Hall, 56 M.J. 432, 436 (C.A.A.F.2002); United States v. Grijalva, 55 M.J. 223, 228 (C.A.A.F.2001). For that reason, we can conduct an independent review of the impact of this constitutional error. See infra, Section II.
b. Does the majority decision, of the Court of Criminal Appeals impeach itself with internally inconsistent statements as to whether denial of a mitigation specialist was harmless ?
The Government’s argument in support of this contention centers on what appears to be a facial inconsistency in the text of the Court of Criminal Appeals’ opinion. As noted, that court concluded that the Government did not meet its burden of showing that the error in denying a mitigation specialist was harmless beyond a reasonable doubt. Kreutzer, 59 M.J. at 779-80. In two footnotes, however, the opinion discusses the test for ineffective assistance of counsel under Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). In that context, Judge Clevenger, the author judge, states:
I specifically do not agree that the prejudice prong of Strickland ... has been satisfied [as to contested issues in the findings]. Notwithstanding the powerful evidence that raises substantial concerns about the quality of Appellant’s mental health, there is still substantial expert opinion evidence of his ability to premeditate and significant direct and circumstantial evidence of the actual process of his alleged premeditation that fact-finders at trial could credit.
Kreutzer, 59 M.J. at 780 n. 8 (citation omitted). Judge Clevenger repeats his conclusion later:
To reiterate, even assuming the first prong of the Strickland test for ineffective assistance of counsel were satisfied, I think [300]*300there was not a reasonable probability that any showing of Appellant’s complete mental health status would have overcome the expert opinion evidence of sanity and the direct and circumstantial evidence of premeditation.
Id. at 784 n. 11 (citation omitted).
The Government argues that with this language the Court of Criminal Appeals impeached its own decision. In the Government’s view Judge Clevenger found that the denial of a mitigation specialist was not harmless beyond a reasonable doubt but also said that the evidence a mitigation specialist would produce had no reasonable probability of changing the result on the findings. The Government argues that this internal inconsistency demonstrates that denial of a mitigation specialist was harmless beyond a reasonable doubt.
Rreutzer disputes the contention that the Army court impeached its own decision by referencing the prejudice test for ineffective assistance of counsel. Kreutzer points out that the test for ineffective assistance involves a different standard and burden when reviewing the effect of denial of competent counsel. He notes that there is no reason to find any inconsistency because it is wholly proper to come to distinct conclusions under two separate tests with the burden falling upon different parties.
We agree that there is an appearance of inconsistency. We do not agree, however, that that inconsistency is necessarily erroneous nor do we find that Judge Clevenger impeached his own decision in his footnotes. Trial and appellate practices are replete with different burdens of proof and persuasion that are allocated to one party or another.
The party benefiting from a constitutional error bears the burden of demonstrating that the error was harmless beyond a reasonable doubt. Chapman, 386 U.S. at 24, 87 S.Ct. 824; Simmons, 59 M.J. at 489. See also Kaiser, 58 M.J. at 149 (citation omitted) (“Error of constitutional dimensions requires either automatic reversal or an inquiry into whether, beyond a reasonable doubt, the error did not contribute to the defendant’s convietion[.]”); Grijalva, 55 M.J. at 228 (“When there has been an error of constitutional dimension, this Court may not affirm unless it is satisfied that the error is harmless beyond a reasonable doubt.”). Thus, in this case, the Government is called upon to show that the error had no causal effect upon the findings. Specifically, the Government had to demonstrate that there was no reasonable possibility that the absence of a mitigation specialist contributed to the contested findings of guilty. See Gutierrez v. McGinnis, 389 F.3d 300, 307-08 (2d Cir.2004).
The test and burden relating to prejudice from ineffective assistance of counsel are different. If an appellant establishes that counsel was ineffective under the first prong of Strickland, it is the appellant as opposed to any party benefiting from error (in this case, the Government) who bears the burden of demonstrating prejudice. United States v. Davis, 60 M.J. 469, 473 (C.A.A.F.2005); United States v. Dewrell, 55 M.J. 131, 133 (C.A.A.F.2001). To establish prejudice, the appellant must demonstrate a reasonable probability that, but for counsel’s deficiency, the result would have been different. Davis (citing United States v. Quick, 59 M.J. 383, 387 (C.A.A.F.2004)). The appellant must demonstrate such prejudice as to indicate a denial of a “fair trial, a trial whose result is unreliable.” Dewrell, 55 M.J. at 133 (citing Strickland, 466 U.S. at 687, 104 S.Ct. 2052). In this setting, overwhelming evidence of guilt may present an insurmountable obstacle to an appellant claiming prejudice from ineffective assistance of counsel.
The tests for determining constitutional harmless error and for determining prejudice under an ineffective assistance analysis are substantially different: the burden falls on different parties (the Government vs. the appellant); the burdens themselves are different (possibility vs. probability); and different considerations are given to the quality and weight of the evidence of guilt in each test. In applying the two tests, it is therefore not unreason[301]*301able or illogical to come to two different conclusions, even in a single case. In light of this, and according Judge Clevenger the presumption that he knew the law, we find no impeaching inconsistency within the opinion. See United, States v. Eversole, 53 M.J. 132, 138 (C.A.A.F.2000).
c. Is the nature of the mental health evidence potentially gathered by a mitigation specialist relevant to the contested elements of premeditation?
The Government acknowledges that general mental health evidence is relevant to establish mens rea, but argues that in this case, due to Kreutzer’s specific personality disorders, any potential mental health evidence that a mitigation specialist could have obtained would not be relevant to the findings. The Government characterizes Kreutzer’s mental health history as demonstrating diminished capacity and asserts that because evidence of diminished capacity is not relevant to the issue of mental responsibility, it could not negate the element of premeditation. Kreutzer, on the other hand, claims that a mitigation specialist would have identified additional mental health evidence as well as helped defense counsel identify and better use experts. Kreutzer asserts that in the end a mitigation specialist would have substantially contributed to presenting a more coherent and stronger mental health theory at trial and that it is possible the members, or at least one member, could have come to a different conclusion on findings.
Under the circumstances of this case we disagree with the Government’s generalization that none of the mental health evidence that was available could or would have an impact upon a member’s determination of premeditation. We have not limited military justice jurisprudence to a narrow use of mental health evidence. Indeed, Ellis v. Jacob, 26 M.J. 90, 93 (C.M.A.1988), dispelled any construction of Article 50a(a), UCMJ, 10 U.S.C. § 850a(a) (2000), that would eliminate evidence of mental conditions relevant to premeditation, specific intent, knowledge, or willfulness, i.e., elements of offenses. See also United States v. Schap, 49 M. J. 317, 322 (C.A.A.F.1998).
The record reflects a wealth of mental health information in this case. A mix of psychological and psychiatric professionals were involved with Kreutzer both before and after the offenses. One mental health professional, Dr. (Colonel) Robert Brown Sr., opined that Kreutzer was “chronically and seriously mentally ill.” This particular information was not known to Kreutzer’s defense counsel prior to trial. Properly prepared and presented, testimony of this nature arguably could go beyond demonstrating diminished capacity and be a substantial part of a defense against the premeditation element. As Judge Clevenger’s lead opinion for the Court of Criminal Appeals points out:
“The mitigation specialist’s role would be to gather and collate appellant’s civilian and military history with a view to the psychiatric issues that would help explain appellant’s state of mind at the critical time of the shooting. One could speculate endlessly on what such an expert, if provided, would have done to help the detailed defense counsel assess the whole story____”
Kreutzer, 59 M.J. at 778-79. Amidst the “wealth of relevant information available to discover, investigate, preserve, analyze, evaluate and potentially exploit at trial in defense of the premeditation allegations,” Judge Cle-venger found “most telling” the fact that Dr. Brown’s “potentially powerful, exculpatory mental status evidence was not discovered by, or known to the defense counsel, at the time of trial.” Id. at 779.
The Government must show there is no reasonable possibility that even a single court member might have harbored a reasonable doubt in light of the mental health evidence that the mitigation specialist could have gathered, analyzed, and assisted the defense to present. Had but a single member harbored a reasonable doubt, death would have been excluded as a permissible punishment. The Government has not met its burden of demonstrating that a mitigation specialist could [302]*302have done nothing to assist counsel to present additional evidence of Kreutzer’s mental health that would not have had an impact on the premeditation element for at least one court member.
II.
De novo review for harmlessness beyond, a reasonable doubt
We have held in this opinion that the Court of Criminal Appeals applied an incorrect definition to the nature of the constitutional harmless error inquiry and that we review de novo the impact of that error in this case. Kreutzer urges us to affirm the Court of Criminal Appeals regardless of the error. He argues that if the Government could not meet the erroneous lower standard applied by the Army court, then it surely could not demonstrate that the error in denying the mitigation specialist was harmless beyond a reasonable doubt.
The Government must demonstrate there is no reasonable possibility that the absence of a mitigation specialist contributed to the contested findings of guilty or, in this case, that not even a single member would have harbored a reasonable doubt after considering the mental health evidence that the mitigation specialist could have gathered, analyzed, and assisted the defense in presenting. We do not believe that the Government has met that burden.
To place this discussion in its proper context, it is necessary to examine the role of a mitigation specialist in capital litigation, both generally and in this case. The general role of a mitigation specialist was discussed in a report adopted by the Judicial Conference of the United States:
Mitigation specialists typically have graduate degrees, such as a Ph.D. or masters degree in social work, and have extensive training and experience in the defense of capital cases. They are generally hired to coordinate an investigation of the defendant’s life history, identify issues requiring evaluation by psychologists, psychiatrists or other medical professionals, and assist attorneys in locating experts and providing documentary material for them to review.9
The American Bar Association recommends the inclusion of a mitigation specialist on every capital litigation defense team and identifies the mitigation specialist as a “core member” of the defense team:
A mitigation specialist is also an indispensable member of the defense team throughout all capital proceedings. Mitigation specialists possess clinical and information-gathering skills and training that most lawyers simply do not have. They have the time and the ability to elicit sensitive, embarrassing and often humiliating evidence (e.g., family sexual abuse) that the defendant may have never disclosed. They have the clinical skills to recognize such things as congenital, mental or neurological conditions, to understand how these conditions may have affected the defendant’s development and behavior, and to identify the most appropriate experts to examine the defendant or testify on his behalf. Moreover, they may be critical to assuring that the client obtains therapeutic services that render him cognitively and emotionally competent to make sound decisions concerning his ease.10
When Kreutzer’s defense attorneys made their requests for a mitigation specialist they supported it with an affidavit from Dr. Lee Norton, Ph.D., a mitigation specialist. Doctor Norton provided extensive background on what a mitigation specialist could provide in regard to mental health evidence.11 In addi[303]*303tion to the general importance of a mitigation specialist in death penalty eases, mitigation specialists may play a particularly important role in ensuring the fair and full adjudication of military death penalty cases where, as here, counsel have little training or experience in capital litigation.
This case is replete with evidence or information indicating that Kreutzer’s mental health was dubious. Yet the presentation of the defense ease-in-chief includes testimony from only three individuals about Kreutzer’s performance, behavior and reputation, and expert testimony from a single mental health professional. Color Sergeant David Wake-land 12 testified that Kreutzer was an average or above-average soldier and noncommis-sioned officer. He also discussed some matters that may have been stressful for Kreut-zer as well as how Kreutzer’s threats and absence were reported on the morning of the shooting.
Specialist Robert Harlan, Kreutzer’s roommate, testified about the ridicule Kreutzer endured, an incident involving threats while the unit was in the Sinai, the chain of command’s treatment of Kreutzer, and Kreut-zer’s emotional state on the day before the shooting. A stipulation of the expected testimony of Sergeant Arthur Swartz provided a further indication that Kreutzer was not respected in the unit.
The only mental health professional called by the defense on the merits was Doctor (Major) Carroll J. Diebold, the Chief of the Department of Psychiatry and Neurology at Womack Army Medical Center, Fort Bragg, North Carolina.13 Accepted as an expert in the fields of psychiatry and forensic psychiatry, Dr. Diebold participated in a sanity board evaluation of Kreutzer in December 1995. He described how the sanity board had proceeded and testified that the board found Kreutzer to have “an adjustment disorder with mixed anxiety and depressed mood” and “dysthymia,” which is “low grade depression.” Generally, Dr. Diebold testified that a person with these personality traits would react to stress with a “spike” into a “detrimental behavior zone.” Finally Dr. Diebold opined that factors similar to those Kreutzer experienced prior to the shooting “would produce a narcissistic injury and an interpretation by Sergeant Kreutzer as a threat on his own character” and cause the “spike” effect.
Trial counsel noted in closing that Dr. Diebold said Kreutzer’s mental disorders “are not even considered severe” and contrasted that with defense counsel’s opening statement promising to show that Kreutzer “was suffering from a severe personality disorder.” Trial counsel also reminded the members that Dr. Diebold testified that none of Kreutzer’s disorders impaired the formation of the intent to kill or the ability to premeditate. Kreutzer’s defense counsel responded that Kreutzer’s actions before the shooting were a “cry for help.” He argued that when the unit started moving out for the run, Kreutzer “spiked” due to the stress and formed the intent to pull the trigger, but that he never intended to kill, nor did he premeditate.
Despite the rather limited defense mental health presentation on the merits, the record on appeal reveals significant additional mental health evidence potentially available to the defense. Doctor Darren Fong, the first mental health professional with whom Kreut-zer spoke did not testify. Kreutzer spoke with Dr. Fong about killing members of his [304]*304unit. However, rather than believing that Kreutzer’s homicidal ideation was serious, Dr. Fong believed that Kreutzer had “problems with anger and very poor coping skills,” “interpersonal problems,” and “probably a history of psychological problems.”
After he was apprehended, Kreutzer saw Dr. (then-Captain) Wendi T. Diamond, the division psychiatrist for the 82d Airborne Division at Fort Bragg. Doctor Diamond indicated that “[n]ever in all my life had I seen someone in so much psychic distress.” She believed that Kreutzer “was not at all rational during our conversation” and that his thoughts were “disordered.” Doctor Diamond was not contacted by the trial or defense counsel prior to trial, a fact that surprised her because she “believed that both sides could have benefited from [her] assessment of SGT Kreutzer’s mental state very close to the time of the offenses.” The potential value of Kreutzer’s discussions with Dr. Diamond to the defense is underscored by the comments of a Government lawyer who observed that interview: “Conclusion: Prepare for Insanity Defense! This guy is nutty [sic] than a fruit cake.” Kreutzer, 59 M.J. at 777.
In pretrial confinement Kreutzer was seen for a suicide assessment by Lieutenant Commander Drew Messer who received his professional training in a dual degree law and psychology program. Lieutenant Commander Messer concluded that Kreutzer was “profoundly depressed” and felt that “there were definite mental health issues in the case.” Nonetheless, he never met with any of Kreutzer’s defense attorneys.
At the defense’s request, mental health professionals at the Walter Reed Army Medical Center Forensic Psychiatric Program were appointed as expert consultants to the defense. Doctor Robert S. Brown Sr. was a consultant to the Forensic Psychiatric Program and participated in evaluating Kreut-zer’s mental condition. Doctor Brown met with Kreutzer on April 11, 1996 and delivered a written report to Dr. Lande, Chief of the Forensic Psychiatric Program, on that same day. Doctor Brown’s report opined that Kreutzer was “chronically and seriously mentally ill” and that the offenses were causally related to his mental illness. While defense counsel did have discussions with Dr. Lande, they were not aware of Dr. Brown or his written evaluation until after trial.
A post-trial mitigation specialist’s report was attached as part of the appellate record. This report reveals that Kreutzer suffered mental and emotional problems prior to entering the Army. He lacked self-confidence and held himself in low esteem, feelings that became worse during junior high school. Symptoms of depression began when Kreut-zer was twelve and increased during his high school years when he also experienced suicidal ideation. Kreutzer continued to be depressed and experienced his first homicidal feelings during his college years. His suicidal ideation manifested itself even more during his college years when he pointed loaded weapons at his head several times.
In contrast to this wealth of mental health information and history favorable to a defense presentation, there is mental health evidence that would indicate that Kreutzer premeditated his actions and was mentally responsible. Doctor Rollins, a privately employed forensic psychologist opined that Kreutzer had no severe mental disorder and was competent to stand trial. Doctor Rollins was not employed by the defense to serve as an expert consultant or witness because of financial considerations, but he recommended that the defense put its main effort into the case in mitigation. Further, a sanity board found Kreutzer competent and not to be suffering from a severe mental disease or defect.
The Government also points to the “interim report” prepared by a mitigation specialist on behalf of the defense for appellate litigation. The Government notes that this appellate mitigation specialist’s report considered the foregoing evidence pertaining to Kreutzer’s mental health and makes no conclusion of lack of mental responsibility or health that would overcome premeditation. Therefore, the Government argues that the mitigation specialist, if provided at trial, would have been relevant only to sentencing and not to findings, and that any error is [305]*305therefore harmless beyond a reasonable doubt.
We are not persuaded that the Government has met its burden of showing that Kreutzer could not have possibly benefited from the talent and expertise of a mitigation specialist on findings. We need not speculate on precisely how the wealth of mental health evidence could have been used at trial. Although capital cases do not confer a per se right to a mitigation specialist, on a case-by-case basis servicemembers confronted with a capital prosecution are entitled to mitigation specialists where their services would be necessary to the defense team. We believe that the Government gives too little weight to the possible worth of a mitigation specialist in this case. The UCMJ and the R.C.M. assure that the defense counsel has the resources, including expert assistance, to prepare and present the defense. See Article 46, UCMJ; R.C.M. 703. The military accused’s rights in this regard are not dependent upon indigence, nevertheless we agree with the spirit of the Arizona Supreme Court’s statement that “[s]o long as the law permits capital sentencing,” the justice system “must provide adequate resources to enable indigents to defend themselves in a reasonable way.” See State v. Bocharski, 200 Ariz. 50, 22 P.3d 43, 55 (2001).
While the services of a mitigation specialist are commonly used in sentencing, in the appropriate case this expert assistance may be necessary to the defense on findings as well. As the Commentary to ABA Death Penalty Counsel Guideline 4.1 states, the mitigation specialist is an “indispensable member of the defense team throughout all capital proceedings.” Kreutzer’s three uniformed attorneys recognized that they could not gather, analyze, and formulate this mental health evidence; a mitigation specialist could have done so and assisted counsel in identifying qualified mental health experts to present the evidence on both the merits and on sentencing. In turn, the defense on the merits could have incorporated that analysis either to bolster the theory that was used at trial or to create a different theory to contest premeditation on the merits. For example, the defense might have used testimony from Dr. Fong to show that he did not take Kreut-zer’s talk about killing seriously and that Kreutzer had a history of homicidal ideation. The defense could have then argued that the members should discount Kreutzer’s night-before statements to SP4 Mays because they were more homicidal fantasy than premeditation. Alternatively, defense counsel might have argued that the additional mental health information produced by a mitigation specialist demonstrated that Kreutzer was susceptible to stress stimuli and was exhibiting “spiked” behavior as opposed to a premeditated intent in committing his crimes. Further, defense counsel may have used the additional information to attack and cloud the findings of the sanity board and try to suggest that while Kreutzer might have been mentally responsible under the law, he did not have the mental capacity to premeditate his crimes. The question is not whether these arguments are persuasive in the abstract, but rather, in light of the fact that Kreutzer was denied the fair opportunity to make these arguments, whether the Government has shown that the error in denying the defense request for a mitigation specialist was harmless beyond a reasonable doubt.
We answer the certified question in the negative. Erroneous denial of Kreutzer’s request for a mitigation specialist was error of constitutional magnitude. As such, the Government must show there was no reasonable possibility that even a single court member might have harbored a reasonable doubt in light of the mental health evidence that the mitigation specialist could have gathered, analyzed, and assisted the defense to present. Had but a single member harbored a reasonable doubt, death would have been excluded as a permissible punishment. In light of these factors, including the relative experience and training of Kreutzer’s defense counsel in capital litigation and the evidence relating to Kreutzer’s mental health history, we hold that the Government has not met its burden of demonstrating that the error in denying Kreutzer’s request for employment of a mitigation specialist was harmless be[306]*306yond a reasonable doubt as to the contested findings.
DECISION
The certified question is answered in the negative and the decision of the United States Army Court of Criminal Appeals is affirmed.