dissenting.
Nearly three decades of observing and participating in the application of the so-called harmless error analysis persuades me that the courts of appeals, by reciting the Bruton standard but justifying its breach in the name of harmlessness, increasingly are eroding the important constitutional protection afforded by the Bru-ton barrier. Incessantly reminding the prosecution and the trial courts of their duty to prevent Bruton error on the one hand, while excusing its violation on the other hand, builds an ever-increasing floor under this unlawful practice rather than placing a ceiling upon it.
The circumstances of this appeal demand a stronger expression of our commitment to Bruton. Here, the Bruton error results directly from the district court’s failure to hold the prosecution to its representation that it would not introduce any statements implicating Bruton. The consequence of that failure — the admission of numerous reinforcing prejudicial statements — goes beyond harmless.
True, Bruton recognized that “ ‘[a] defendant is entitled .to a fair trial but not a perfect one.’ ” Bruton v. United States, 391 U.S. 123, 135, 88 S.Ct. 1620, 20 L.Ed.2d 476 (1968) (quoting Lutwak v. United States, 344 U.S. 604, 619, 73 S.Ct. 481, 97 L.Ed. 593 (1953)). We no longer adhere to the “time in the law, extending into our own century, when no error was lightly forgiven.” Roger J. Traynor, The Riddle of Harmless Error 3 (1970) (“In that somber age of technicality the slightest error in a trial could spoil the judgment.”). Accordingly, the harmless error test permits the courts of appeals to affirm decisions despite the commission of constitutional errors. Yet, application of this test does not justify shirking our duty to protect the fairness of trials. Our duty entails analyzing the effect of the improper evidence *1222upon the properly admitted evidence. United States v. Glass, 128 F.3d 1398, 1403 (10th Cir.1997) (“To hold an error of constitutional dimension harmless, we must conclude ‘the properly admitted evidence of guilt is so overwhelming, and the prejudicial effect of the codefendant’s admission is so insignificant by comparison, that it is clear beyond a reasonable doubt that the improper use of the admission was harmless error.’ ” (quoting Schneble v. Florida, 405 U.S. 427, 430, 92 S.Ct. 1056, 31 L.Ed.2d 340 (1972)) (emphasis added)).
The majority takes its formulation of the harmless error standard from Washington v. Recuenco, — U.S.-, 126 S.Ct. 2546, 2552, 165 L.Ed.2d 466 (2006), a case in which the Court was not evaluating the harmlessness of the conceded error, but rather alluding to the standard that the state court should have employed. I prefer the more precise definition of the harmless error test provided in Neder v. United States, the ease cited by the Recuenco Court in its discussion of harmless error: “[T]he test for determining whether a constitutional error is harmless ... is whether it appears ‘beyond a reasonable doubt that the error complained of did not contribute to the verdict obtained.’ ” 527 U.S. 1, 16, 119 S.Ct. 1827, 144 L.Ed.2d 35 (1999) (quoting Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967)). Since the test for sufficiency of the evidence to convict is “whether any rational juror” could find guilt beyond a reasonable doubt, Schlup v. Delo, 513 U.S. 298, 330, 115 S.Ct. 851, 130 L.Ed.2d 808 (1995) (reciting standard promulgated in Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979)) (emphasis added), it seems appropriate that the harmless error test is best disciplined by asking the inverse: could any reasonable juror entertain a reasonable doubt that the prejudicial effect of the Bruton error contributed to the jury’s decision? See Sullivan v. Louisiana, 508 U.S. 275, 279, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993) (holding that correct inquiry for harmless error review is “whether the guilty verdict actually rendered in this trial was surely unattributable to the error” (second emphasis added)).
Nevertheless, the majority does not appear to apply even the recited standard. While acknowledging that it must consider “ ‘the probable impact’ ” of the Bruton evidence “ ‘on the minds of an average jury’ ” based upon an analysis of “the context in which the Bruton statement was admitted, how it was used at trial, and how it compares to the properly admitted evidence” (ante at 1219) (quoting and citing United States v. Sarracino, 340 F.3d 1148, 1160 (10th Cir.2003)), the majority foregoes any analysis of these considerations, instead simply concluding that its laundry list of properly admitted evidence — much of it contested at trial — is sufficient to establish Defendant’s guilt. By wholly ignoring the significance of the wrongly admitted Bruton evidence, the majority fails to address the most significant part of the standard.1 “A reviewing court making this harmless-error inquiry does not, as Justice Traynor put it, ‘become in effect a second jury to determine whether the defendant is guilty.’ ” Neder, 527 U.S. at 18, 119 S.Ct. 1827 (quoting Traynor, supra, at 21). Instead, we are meant to determine “what effect the error had or reasonably may be taken to have had upon the jury’s decision,” assessed not based upon a judge’s personal reaction, but “with allowance for how others might react and not be regard*1223ed generally as acting without reason.” Kotteakos v. United States, 328 U.S. 750, 764, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946). Still, the majority discounts without consideration the prejudicial effect that the significant Bruton evidence must have had upon the jury.
As an initial matter, we should assess the importance that the seasoned prosecutors placed upon that wrongly admitted evidence when they looked ahead at what the jury would think of their case, rather than looking back after the jury had resolved the clear disputes in the evidence in their favor. When the accused was doing all he could to keep this evidence out, including moving to sever and moving in limine to preclude admission of these statements, the prosecution vigorously resisted. Indeed, despite stating that it had no intention of introducing Bruton evidence, the prosecution proceeded to introduce a host of prejudicial testimony and relied upon both that testimony and the implications contained therein during its closing argument. See Arizona v. Fulminante, 499 U.S. 279, 297-98, 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991) (noting that prosecution viewed coerced confession as essential and emphasized erroneously admitted confession in opening and closing statements); Satterwhite v. Texas, 486 U.S. 249, 260, 108 S.Ct. 1792, 100 L.Ed.2d 284 (1988) (noting prosecutor’s emphasis of erroneously admitted testimony in capital sentencing proceeding); see also United States v. Sarracino, 340 F.3d 1148, 1164 (10th Cir.2003) (assessing effect of prosecution’s repeated emphasis of improperly admitted statement). Viewing the importance of the Brutm testimony from the prosecution’s perspective, it is apparent that the prosecution believed the Bruton evidence to be essential to its case, which strikes a stark contrast to its claim on appeal that the introduction of this evidence was without consequence.
In assessing whether the district court’s failure to enforce the standards of admissibility mandated by Bruton may have influenced the jury, it is essential to keep in mind that the critical admissible evidence and the inferences to be drawn therefrom were called into question by Mr. Nash’s own sworn denial, a rarity amongst Bruton cases. While his denial would not be significant standing alone, the jury’s implicit credibility determination against Mr. Nash becomes suspect once the trial is stripped of the improperly admitted testimony: the corroborating effect of the impermissible statements made by three government witnesses is muted. See Fulminante, 499 U.S. at 298, 111 S.Ct. 1246 (observing that jury’s assessment of improperly admitted confession “could easily have depended in large part on the presence” of other testimony); see also Neder, 527 U.S. at 18, 119 S.Ct. 1827 (stating that Bruton error “infringe[s] upon the jury’s factfinding role and affect[s] the jury’s deliberative process in ways that are, strictly speaking, not readily calculable”). This effect is especially powerful given defense counsel’s impeachment of these witnesses, who included three inmates — two of them friends— trading testimony for sentence reductions. See Fulminante, 499 U.S. at 300, 111 S.Ct. 1246 (noting jury’s determination might have been swayed by testifying witness’s desire for favorable treatment). In addition, the absence of such bolstering would no doubt impact any rational juror’s determination of the credibility of the government’s cooperating witness, who testified to avoid drug distribution charges.
One of the most damaging pieces of admissible evidence — the claimed confession to a cellmate whose testimony in exchange for important benefits could not be characterized as unimpeachable — was directly denied. In relying upon this evidence, the majority usurps the jury’s role by implicitly determining that even absent *1224the Bruton testimony, Defendant’s testimony remains not credible. See Kotteakos, 328 U.S. at 763, 66 S.Ct. 1239 (“[I]t is not the appellate court’s function to determine guilt or innocence.”); cf. Zappulla v. New York, 391 F.3d 462, 470 n. 3 (2d Cir.2004) (collecting reports indicating inherently unreliable nature of jail house informants). This approach, which turns a blind eye on the bolstering and corroborating effect of the Bruton evidence upon the jury’s credibility assessments, does not in my judgment conform to the limited role of the harmless error analysis when deciding whether to excuse the violation of constitutional standards.
As a result of this error, the majority places undue emphasis on the importance of a host of circumstantial evidence and disregards Mr. Nash’s proffered explanations for the existence of this evidence. See United States v. Hill, 901 F.2d 880, 885 (10th Cir.1990) (ruling Bruton error not harmless where much of government’s admissible evidence was circumstantial and defendant offered explanations countering government’s suggested inferences). The record, while not a model of clarity, reflects that Mr. Nash never discussed the drug transaction with Mr. Jones. Mr. Jones conceded as much both on direct examination and on cross-examination. Mr. Jones admitted that Mr. Nash never left his vehicle on the occasions when Mr. Nash drove Mr. Kinchion to Mr. Jones’ store. Moreover, Mr. Jones’ testimony illustrates that his calls to Mr. Nash’s cell phone were made in order to speak with Mr. Kinchion, who apparently did not have a cell phone of his own. Mr. Jones also testified that Mr. Kinchion first approached him at his store without Mr. Nash, having been driven there by a woman. This contradicts police testimony that Mr. Nash first approached Mr. Jones. Additionally, according to Mr. Nash, the $2,800 found in his possession came almost entirely from the recent sale of two cars, which he was able to establish with sales receipts. Mr. Nash’s father confirmed that Mr. Nash frequently bought and sold cars to make money. Mr. Nash also argued that his name likely appeared on the registration of one of the five collateral vehicles because Mr. Kinchion improperly completed the re-registration after Mr. Nash sold the car to Mr. Kinchion. Also, Mr. Nash was a regular customer of the car rental shop, as acknowledged by the rental car owner, rendering his use of the Grand Am innocuous. Whether Mr. Nash intentionally swerved toward the edge of the bridge in order to facilitate the ejection of drugs by Mr. Kinchion, who had been tossing drugs out of the window well before the car approached the bridge, is not clear from the record. Lastly, the majority infects its own decision with Bru-ton evidence by relying on Mr. Jones’ testimony that Mr. Kinchion claimed Mr. Nash was nervous after seeing a police car parked nearby.
The majority’s failure to address the overall effect of the Bruton evidence strips the harmlessness test of a crucial component while reducing the overwhelming evidence requirement to no more than a sufficiency-of-the-evidence examination. Because I cannot say that the cumulative Bruton error was harmless beyond a reasonable doubt, I would reverse and remand for a new trial.
. Certainly hosts of cases merely rule Bruton error harmless due to overwhelming record evidence. While most are simpler cases, to the extent that these decisions are based solely on the presence or absence of "overwhelming” evidence, they do an injustice to the harmless error standard.