(concurring in part and dissenting in part):
I agree with the majority that trial counsel erroneously invited the members to compare the evidence presented on each offense to find propensity. Had the trial counsel desired to make that argument, he should have followed the procedural steps of Military Rule of Evidence (M.R.E.) 413(b). Had those procedural steps been followed, the military judge would have made the necessary threshold findings under M.R.E. 4131 and would have conducted an M.R.E. 403 balancing analysis. Because trial counsel did not comply with the steps for presenting or arguing propensity evidence, the military judge did not evaluate the evidence for admissibility as propensity evidence. Therefore, trial counsel erred by invoking propensity in his argument. I do not agree that the risks created by the improper argument were properly addressed by the spillover instruction. However, I need not determine whether trial counsel’s error was a plain error requiring relief because I conclude that the military judge committed plain error by failing to provide a propensity instruction to the members.
Before closing arguments, the military judge’s instructions to the members included a standard “spillover” instruction. Specifically, the military judge stated, “[I]f you find or believe that the accused is guilty of one offense, you may not use that finding or belief as a basis for inferring, assuming, or proving that he committed any other offense.” Despite this instruction, trial counsel urged the members in his closing argument to compare the offenses because “[i]t will also show you that he has [the] propensity to engage in this sort of conduct.” (emphasis added). Trial counsel went on to urge that consideration of this propensity evidence would not conflict with the military judge’s spillover instruction:
*157Now, before I get to [a comparison of the similarities between the two alleged sexual assaults] I want to preface — this was something the judge told you, I don’t intend for you to take proof of one offense to find him guilty of another, the judge told you that you can’t do that. But what you can do is you can take these things and compare them for his propensity to commit these types of offenses. That’s perfectly acceptable when you’re deliberating.
Emphasis added.
Not only was trial counsel’s invitation to compare the offenses for propensity in direct conflict with the spillover instruction given by the military judge, he erroneously explained to the members that they could consider the propensity evidence despite the spillover instruction. The military judge should have corrected that conflict sua sponte by providing a propensity instruction.
Propensity evidence may be considered by the members to prove a charged substantive offense of sexual assault when the procedures of M.R.E. 413 have been followed. See United States v. Schroder, 65 M.J. 49, 52 (C.A.A.F.2007). However, even when the procedures of M.R.E. 413 have been complied with, this court has further held that the procedural safeguards “required to protect the accused from unconstitutional application of M.R.E. 413 ... include the requirement of proper instructions.” Id. at 55.
Without deciding whether the trial counsel’s propensity argument constituted plain error, absent an instruction as to the proper consideration of propensity evidence, the members had no guidelines as to how to resolve the conflict between the military judge’s instruction and trial counsel’s argument that the instruction could be ignored in this situation.2 The propensity instruction was necessary to prevent the members from convicting Burton on the basis of other than direct evidence of the charged offense and to preclude “reliev[ing] the government of its constitutional burden to prove every element of the charged offense beyond a reasonable doubt.” See id.3 Failure to instruct the members how they should properly consider propensity to commit sexual assault was, under the circumstances of this case, error that was plain and obvious.
In light of the fundamental, constitutional nature of this error, the Government has the burden of establishing that the error had “no causal effect upon the findings.” United States v. Othuru, 65 M.J. 375, 377 (C.A.A.F. 2007) (citing United States v. Simmons, 59 M.J. 485, 489 (C.A.A.F.2004); United States v. Bins, 43 M.J. 79, 86 (C.A.A.F.1995)). The Government must demonstrate that there is no reasonable possibility that the lack of instruction contributed to the contested findings of guilty. United States v. Kreutzer, 61 M.J. 293, 300 (C.A.A.F.2005). Because the members lacked any guidance in the evaluation of trial counsel’s invitation to consider propensity, there is no assurance that the Government was held to its burden of proof or that Burton was convicted on direct evidence of the charged rape rather than upon the improper use of propensity derived from the other charged offense. The error was not harmless beyond a reasonable doubt.
I would reverse the decision of the United States Air Force Court of Criminal Appeals, set aside the findings and sentence, and authorize a rehearing.
. Those required findings are that: "(1) [t]he accused is charged with an offense of sexual assault” defined by M.R.E. 413(d); (2) ‘‘[t]he evidence proffered is 'evidence of the defendant’s commission of another offense of ... sexual assault'; and (3)[t]he evidence is relevant under [M.R.E.] 401 and [M.R.E.] 402.” United States v. Wright, 53 M.J. 476, 482 (C.A.A.F.2000) (quoting United States v. Guardia, 135 F.3d 1326, 1328 (10th Cir.1998)) (requiring threshold findings before admitting evidence under M.R.E. 413); see also United States v. Dewrell, 55 M.J. 131, 138 n. 4 (C.A.A.F.2001).
. For an example of a propensity instruction, see Dep’t of the Army, Pam. 27-9, Legal Services, Military Judges’ Benchbook ch. 7, para. 7-13-1, n. 4 (2002).
. As this court said in Schroder, "[I]t is essential that ... the members ... be instructed that the introduction of such propensity evidence does not relieve the government of its burden of proving every element of every offense charged. Moreover, the factfinder may not convict on the basis of propensity evidence alone.” Id. at 56.