dissenting.
Because the majority ignores United, States v. Bell, 516 F.3d 432 (6th Cir.2008), and misapplies United States v. Jenkins, 593 F.3d 480 (6th Cir.2010), to reach a conclusion that the record and the Federal Rules of Evidence do not support, I respectfully dissent.
Bell’s holding that Rule 404(b) evidence is probative of intent only “when the prior distributions were part of the same scheme or involved a similar modus operandi as the present offense” is not at odds with our prior case law and must be followed. Bell, 516 F.3d at 443. The majority disagrees and points to a series of cases which they claim preclude Bell’s finding on probity; none of them do. The 404(b) evidence admitted in United States v. Myers, 123 F.3d 350, 354 (6th Cir.1997), was probative of intent because it was “part of the same scheme.” Bell, 516 F.3d at 443. All of the other acts in that case occurred during the period where Myers and a co-conspirator were engaged in a scheme to sell crack cocaine. See Myers, 123 F.3d at 354. United States v. Ayoub, 498 F.3d 532 (6th Cir.2007), is of no moment either because Ayoub never contested probative value, so Ayoub’s discussion of the issue is dicta. See Ayoub, 498 F.3d at 548. And United States v. Mauldin, 109 F.3d 1159 (6th Cir.1997), does not bar Bell’s holding because Mauldin did not fully describe what the other-acts evidence consisted of or how it related to the charged crime, and the little that it did say about that issue is not inconsistent with a finding that the prior-acts reflected the same scheme or modus operandi as the charged conduct. See Mauldin, 109 F.3d at 1161.
Applying Bell to the case at hand bars the majority’s finding on probity. The other acts admitted by the district court were not probative of intent because they did not form part of the same scheme or show a similar modus operandi — a point underscored by the government’s failure to argue in the alternative that they did.
The majority’s Rule 403 balancing analysis is also off the mark. I see no way to distinguish this case from Jenkins, which vacated the defendant’s conviction after finding that the other-acts evidence was more prejudicial than probative. Just as in Jenkins, intent was nominally at issue in Hardy’s trial but possession was “the battleground.” Id. at 486. Also like in Jenkins, the government’s evidence of intent to distribute here was overwhelming' — a detective testified that the quantity of cocaine discovered at the house (along with a digital scale) could satisfy 1,500 crack ad-*160diets, and Hardy’s cellmate said, in essence, that Hardy told him that his entire defense was “a trick”- — so the probative value of the other-acts evidence was “microscopic at best” and admitting it was plainly “piling on.” Id. at 485-86. The prejudicial effect of the evidence, moreover, was obvious — once a drug dealer, always a drug dealer — and made worse by the fact that, unlike in Jenkins, the evidence of prior drug-dealing had not been found by a jury beyond a reasonable doubt.
The harmlessness analysis tracks Jenkins too. The error at trial cannot be ignored because the government’s evidence of constructive possession — a separate element — “though solid, was not overwhelming.” Id. at 486. I therefore cannot “say with fair assurance that the admission of the [404(b) evidence] ... did not substantially sway the results at trial.” Id. (internal quotation marks omitted). Moreover, Ayoub does not undermine the analogy between this case and Jenkins. The defendant in Ayoub conceded that the evidence on all the elements of drug distribution was overwhelming, unlike here, where the defendant maintains — and I agree— that the evidence of possession was not so strong that the improper admission of Hardy’s prior bad acts can be said not to have prejudiced the jury’s consideration of that element.
The majority insists that I have the cases wrong and that Bell cannot be reconciled with what came before it. Not so. Bell did not overrule any prior holdings. In finding that prior evidence of drug distribution is probative of intent only where “the prior distributions were part of the same scheme or involved a similar modus operandi as the present offense” Bell gave a precise shape to what other panels has passed on sub silentio; a perfectly legitimate — and binding — exercise of panel authority. Bell, 516 F.3d at 444; see Hilliard v. United States, 157 F.3d 444, 449 (6th Cir.1998) (describing subsequent panel’s refinement of a prior panel rule); cf. United States v. Bowers, 615 F.3d 715, 721-22 (6th Cir.2010) (quoting Hagans v. Lavine, 415 U.S. 528, 533 n. 5, 94 S.Ct. 1372, 39 L.Ed.2d 577 (1974) (“[W]hen questions of jurisdiction have been passed on in prior decisions sub silentio, [we are not] bound when a subsequent case finally brings the jurisdictional issue before us.”)). Viewed this way, Bell’s “scheme or modus operandi” test is not a break with the longstanding requirement that evidence of prior bad acts are probative only where they are “substantially similar and reasonably near in time” to the charges in the indictment, but a refinement of it as applied to drug distribution cases. United States v. Haywood, 280 F.3d 715, 721 (6th Cir.2002) (quoting United States v. Blankenship, 775 F.2d 735, 739 (6th Cir.1985)).
The modesty and propriety of Bell’s holding on probity is evident on considering the definitions of “scheme” and “modus operandi.” The latter means “[a] method of operating or a manner of procedure,” 1 Black’s Law Dictionary (9th ed. Online Ver.2009), and the former “a plan of action devised in order to attain some end.” Oxford English Dictionary, Def. 5.b. (Online Ver. Mar. 2011). These broad definitions plainly encompass the facts of our prior holdings on probity while giving more nuanced guidance to future courts tasked with applying our rule that other acts evidence must be “substantially similar and reasonably near in time” to be probative. Haywood, 280 F.3d at 721.
Consider Ayoub: the other-acts evidence there consisted of testimony describ*161ing a search of Ayoub’s residence in 2000 which yielded “approximately three pounds of marijuana in gallon-size bags, approximately five grams of cocaine, a triple-beam scale and a Ruger .45-caliber handgun.” Ayoub, 498 F.3d at 547. Likewise, scales, handguns, plastic bags, and a pound of marijuana turned up in the search of Ayoub’s parents’ residence in 2004, which led ultimately to the federal conviction we reviewed in Ayoub. See id. at 536. To be sure, this evidence from 2000 related to conduct that is “substantially similar” to the conduct adjudicated at Ayoub’s 2005 trial, but the comparison also reveals a strikingly parallel scheme or modus operandi to distribute marijuana.
As for United States v. Johnson, 27 F.3d 1186 (6th Cir.1994), it is not in tension with Bell either. The majority sees a conflict between the Johnson court’s holding that “where the crime charged is one requiring specific intent, the prosecutor may use 404(b) evidence to prove that the defendant acted with the specific intent” and Bell’s scheme/modus operandi rule. But this passage misleads by making it appear as if the Johnson court held that 404(b) evidence — of any sort, and irrespective of its similarity to the charged conduct — is always probative where specific intent is an element of the charged crime. As the majority obliquely acknowledges, however, Johnson actually found, consistent with prior precedent, that “similar acts of possession with intent to distribute” were probative of the specific intent at issue in that case. Johnson, 27 F.3d at 1193 (emphasis added). So the question remains whether Bell legitimately refined what “similar” means for drug-distribution crimes by looking closely at the prior case law and expressing in more specific language our prior sub silentio holdings. I believe it did and the facts of Johnson underscore further that prior panels have been applying without saying so, the schem e/modus operandi rule that Bell properly articulated.
And, as with our other cases, the facts in Johnson conform to Bell’s rule. There, the evidence of prior bad acts consisted of testimony by a confidential informant that Johnson had sold him crack cocaine on two occasions one month prior to the unconsummated sale of crack cocaine (apparently orchestrated by the same informant) that led to the charging of the crime of indictment. See Johnson, 27 F.3d at 1189. While the opinion does not make clear whether the prior acts were undertaken with the same modus operandi as the charged conduct, it is plain that the prior acts were part of the same scheme to sell drugs as that charged in the crime of indictment, that is, the same “plan of action devised in order to attain [the same] end.” Oxford English Dictionary, Def. 5.b.
Finally, another word on balancing. Jenkins aptly described prior bad acts evidence as “loom[ing] Kong-like” at trial. Jenkins, 593 F.3d at 486. Such evidence dominates and overwhelms every decision a jury makes, and instructions intended to mitigate this effect are fig-leaves that cannot hope to hide the propensity evidence glaring into the jury box. Id. This view of 404(b) evidence is grounded in more than the musings of appeals court judges, empirical studies confirm that “juries treat prior bad acts evidence as highly probative of the charged crime.” United States v. Amayar-Manzanares, 377 F.3d 39, 49 (1st Cir.2004) (Torruella, J., dissenting) (citing Harry Kalven, Jr. & Hans Zeisel, The American Jury 160 (1966); Abraham P. Ordover, Balancing the Presumptions of Guilt and Innocence: Rules 101(b), 608(b) and 609(a), 38 Emory L.J. 135 (1989); Roselle L. Wissler & Michael J. Saks, On the Inefficacy of Limiting Instructions: When Jurors Use Prior Conviction Evidence To Decide on Guilt, 9 L. & Hum. Behav. 37 (1985)). Of course, the real-*162world effects of 404(b) evidence do not necessarily dictate the result in a particular case. Whatever the academy says, we have approved of limiting instructions, and other-acts evidence is sometimes admissible despite its prejudicial effect. But I cannot agree that the district court’s determination here is entitled to our deference.
The majority credits the district court’s finding that the probative value of the other-acts evidence outweighed its prejudicial effect because the government did not have evidence of actual possession. Yet if the prior bad acts admitted below were probative at all they were probative of intent, not actual possession. Thus, the district court’s effort at balancing fails on its own terms, leaving the 404(b) testimony to poison the jury’s consideration of possession, an element on which the government’s evidence, while strong, was not overwhelming. The error, then, was far from harmless.
For these reasons, I would REVERSE the district court’s 404(b) determination and REMAND for a new trial.
. Black’s also describes the more specific— and restrictive — sense: “a patter of criminal behavior so distinctive that investigators attribute it to the work of the same person.” Black’s Law Dictionary (9th ed. Online Ver. 2009).