United States v. Rafael Santana and Francis Fuentes

SELYA, Circuit Judge. In the six decades since Justice Roberts noted that “[s]ociety is at war with the criminal classes,” Sorrells v. United States, 287 U.S. 435, 453, 53 S.Ct. 210, 216, 77 L.Ed. 413 (1932) (Roberts, J., dissenting), hostilities have escalated and armaments have grown more destructive. Here, the government’s weapon was 13.3 grams of heroin, 92% pure, delivered into the stream of commerce as part of an effort to gain the confidence of suspected drug traffickers. The district judge decided that the government’s guerilla tactics impermissibly endangered civilians and dismissed the ensuing charge. See United States v. Santana, 808 F.Supp. 77 (D.Mass.1992). The United States appeals. Although law enforcement officers might well profit from reading the lower court’s thoughtful opinion, we conclude that the court exceeded its authority. Consequently, we reverse. I. BACKGROUND In 1991, the federal Drug Enforcement Administration (DEA) mounted an elaborate reverse sting designed to bring a mammoth heroin distribution network to ground. The DEA believed that defendant-appellee Rafael Santana ran the ring from prison through various henchmen, including defendant-ap-pellee Francis Fuentes. In the course of the sting, Fuentes asked an undercover agent, posing as a heroin supplier, to furnish a sample of his wares. The agent received a special dispensation from DEA hierarchs and delivered 13.3 grams of heroin, 92% pure, to Fuentes in August of 1991. 1 The authorities never recovered the sample. There is a factual dispute over the size of the stakes. The government, based on its agent’s testimony, claims that the deal under negotiation contemplated delivery of 141 kilograms of heroin. It further claims, based on an informer’s account, that Santana’s organization was capable of distributing up to 200 kilograms of heroin monthly. Appellees suggest that the negotiations concerned a considerably smaller quantity of narcotics, and that the organization, if it existed at all, was far less ambitious. We need not enter this thicket; for present purposes, the relevant finding is the reasonableness, at the time the sample was furnished, of the government’s belief that the alleged organization had the capacity to manage widespread distribution of heroin. It is not seriously disputed that the government thought this to be the ease; *3 and, moreover, the government’s belief, given both the information in its files and Santana’s history — he had been convicted in 1990 of conspiracy to smuggle 1,000 kilograms of heroin — was objectively reasonable. Having been made privy to the evidence collected in the course of the government’s indagation, a federal grand jury returned a three-count indictment against seven defendants, including appellees, in October of 1991. The defendants filed pretrial motions seeking to dismiss the indictment on the ground that the government acted outrageously in fronting so much heroin and then losing track of it. A magistrate judge recommended that the motions be denied. The district court rejected the recommendation. Presuming that most of the unretrieved sample reached end users, see id. at 79, the court found that the government’s actions exceeded the bounds of propriety, see id. at 81-84. It thereupon dismissed count 3 of the indictment (the count for which the 13.3-gram sample formed the corpus delicti). 2 See id. at 85-86. The court derived its authority from the due process clause of the Fifth Amendment, and, alternatively, from its supervisory power. See id. at 86. The government moved unsuccessfully for reconsideration 3 and now appeals. We have jurisdiction under 18 U.S.C. § 3731 (1988). There are two main legal points in contention. First, the government denies that its conduct was outrageous. Second, the government asseverates that the district court lacked authority under either the due process clause or the rubric of supervisory power to redress injuries to third parties by dismissing charges against appellees. On the facts of this case, we think that both points are well taken. II. THE DOCTRINE OF OUTRAGEOUS MISCONDUCT Outrageous misconduct is the deathbed child of objective entrapment, a doctrine long since discarded in the federal courts. See, e.g., Sherman v. United States, 356 U.S. 369, 372, 78 S.Ct. 819, 820, 2 L.Ed.2d 848 (1958) (rejecting an objective entrapment approach in favor of a subjective approach). The doctrine’s midwife was Chief Justice Rehnquist (then Justice Rehnquist), who, in the course of championing a subjective theory of entrapment, speculated that the Court might “some day be presented with a situation in which the conduct of law enforcement agents is so outrageous that due process principles would absolutely bar the government from invoking judicial processes to obtain a conviction.... ” United States v. Russell, 411 U.S. 423, 431-32, 93 S.Ct. 1637, 1643, 36 L.Ed.2d 366 (1973). Seizing upon this dictum, the defendant in Hampton v. United States, 425 U.S. 484, 96 S.Ct. 1646, 48 L.Ed.2d 113 (1976), attempted to construct an outrageous misconduct defense rooted in the due process clause. Hampton lost his case but succeeded in legitimating the doctrine, albeit precariously. 4 Although it has a comfortably familiar ring, “outrageous misconduct” is surpassingly difficult to translate into a closely defined set of behavioral norms. The broadest hints as to the content of the outrageousness standard lie in the dictum that spawned the doctrine. Inasmuch as Rochin v. California, 342 U.S. 165, 72 S.Ct. 205, 96 L.Ed. 183 (1952), is the case irrefragably linked with *4 the legal rubric of fundamental fairness, one hint is found in Justice Rehnquist’s citation to Rochin. See Russell, 411 U.S. at 431-32, 93 S.Ct. at 1642-43. A second hint is contained in Russell’s explicit equation of outrageous misconduct with violations of “that ‘fundamental fairness, shocking to the universal sense of justice,’ mandated by the Due Process Clause of the Fifth Amendment.” Russell, 411 U.S. at 432, 93 S.Ct. at 1643 (quoting Kinsella v. United States ex rel. Singleton, 361 U.S. 234, 246, 80 S.Ct. 297, 303, 4 L.Ed.2d 268 (1960)). Picking up on these clues, most courts apply a variant on the fundamental fairness standard as a sounding line for outrageousness. See Mosley, 965 F.2d at 910 (collecting formulations). Although this standard lacks mathematical precision, we agree with Justice Frankfurter that imprecision of this nature does not leave courts without adequate guidance; rather, “[i]n dealing not with the machinery of government but with human rights, the absence of formal exactitude, or want of fixity of meaning, is not an unusual or even regrettable attribute of constitutional provisions.” Rochin, 342 U.S. at 169, 72 S.Ct. at 208. The banner of outrageous misconduct is often raised but seldom saluted. Even though one respected jurist contends that the doctrine belongs in the dustbin of history, see United States v. Miller, 891 F.2d 1265, 1271-73 (7th Cir.1989) (Easterbrook, J., concurring), 5 case after case confirms its continued existence. See Moran v. Burbine, 475 U.S. 412, 432, 106 S.Ct. 1135, 1146, 89 L.Ed.2d 410 (1986) (“We do not question that on facts more egregious than those presented here police deception might rise to a level of a due process violation.”); United States v. Mosley, 965 F.2d 906, 909 (10th Cir.1992) (collecting cases from eleven circuits). Be that as it may, the doctrine is moribund; in practice, courts have rejected its application with almost monotonous regularity. See, e.g., United States v. Barnett, 989 F.2d 546, 560 (1st Cir.1993), petition for cert. filed (June 28, 1993) (No. 93-5018); United States v. Lilly, 983 F.2d 300, 309-10 (1st Cir.1992); United States v. Marino, 936 F.2d 23, 27 (1st Cir.1991); United States v. Rosen, 929 F.2d 839, 842 (1st Cir.), cert. denied, — U.S. -, 112 S.Ct. 77, 116 L.Ed.2d 51 (1991); United States v. McDowell, 918 F.2d 1004, 1008-09 (1st Cir.1990); see also United States v. Panitz, 907 F.2d 1267, 1272-73 (1st Cir.1990) (collecting pre-1990 First Circuit cases declining to invoke the doctrine); United States v. Bogart, 783 F.2d 1428, 1434-38 (9th Cir.) (summarizing relevant case law), vacated in part on other grounds sub nom. United States v. Wingender, 790 F.2d 802 (9th Cir.1986); United States v. Warren, 747 F.2d 1339, 1342-43 & nn. 7-8 (10th Cir.1984) (collecting precedents from various circuits). Indeed, since the Supreme Court decided Hampton, a federal appellate court has granted relief to a criminal defendant on the basis of the outrageous misconduct defense only once. See United States v. Twigg, 588 F.2d 373, 382 (3d Cir.1978). The historical record makes it clear, therefore, that the outrageous misconduct defense is almost never successful. 6 There are two competing visions of the doctrine’s role. One school of thought holds that the defense should be confined to cases involving extreme physical, and possibly psychological, abuse of a defendant. See United States v. Kelly, 707 F.2d 1460, 1476 n. 13 (D.C.Cir.) (per curiam) (collecting eases), *5 cert. denied, 464 U.S. 908, 104 S.Ct. 264, 78 L.Ed.2d 247 (1983). A second school of thought holds that outrageous misconduct may also function as a kind of supplement to the entrapment defense, reserved for those cases where law enforcement personnel become so overinvolved in a felonious venture that they can fairly be said either to have “creat[ed]” the crime or to have “coere[ed]” the defendant’s participation in it. Mosley, 965 F.2d at 911-12; see also Bogart, 783 F.2d at 1436-38. This case does not require us to choose between these two conceptions of the doctrine. III. APPLYING THE DOCTRINE Having traced the evolution of the doctrine of outrageous misconduct, we proceed to consider its applicability in this case. Although what transpired here fits neither of the conventional patterns of outrageous misconduct described above, the district court nonetheless ruled that furnishing the hefty heroin sample (and then losing track of it) comes within the doctrine’s sweep. We conclude, for two independently sufficient reasons, that the district court erred. A. Outrageousness. It is incontrovertible that the government may supply drugs to a suspect in the course of a drug investigation. See Hampton, 425 U.S. at 491, 96 S.Ct. at 1650 (Powell, J., concurring). When this occurs in the prototypical case, an agent documents a malefactor’s acceptance of a government-supplied sample and then promptly arrests him. In this scenario, even a large quantity of government-supplied drugs will not raise judicial eyebrows, for the contraband is regained coincident with the arrest. See, e.g., Barnett, 989 F.2d at 560 (declining to find outrageous misconduct where agent sold suspect enough hydriodic acid to manufacture 18 kilos of methamphetamine but recovered it promptly); United States v. Gianni, 678 F.2d 956, 960 (11th Cir.) (similar; agents sold suspect 1150 lbs. of marijuana but recovered it promptly), cert. denied, 459 U.S. 1071, 103 S.Ct. 491, 74 L.Ed.2d 633 (1982); United States v. Dunn, 608 F.Supp. 530, 531 (W.D.N.Y.1985) (similar; agent sold suspect one kilo of cocaine but recovered it promptly). The government’s role in supplying drugs is more problematic when the drugs are not recovered. Nonetheless, several courts have held that providing a known addict small quantities of drugs in order to facilitate the progress of an undercover agent’s work does not constitute outrageous misconduct. See United States v. Harris, 997 F.2d 812, 817-19 (10th Cir.1993); United States v. Barrera-Moreno, 951 F.2d 1089, 1092 (9th Cir.1991), ce rt. denied, — U.S. -, 113 S.Ct. 417, 121 L.Ed.2d 340 (1992) & - U.S. -, 113 S.Ct. 985, 122 L.Ed.2d 137 (1993); United States v. Ford, 918 F.2d 1343, 1349-50 (8th Cir.1990). Common sense suggests that, where the target of the investigation is a distributor rather than an addict, the quantity of drugs needed to earn or retain the suspect’s confidence will likely be larger. 7 It is, therefore, unsurprising that courts generally have declined to find outrageous misconduct in situations of this sort despite the disappearance of fairly substantial quantities of government supplied contraband. See, e.g., United States v. Valona, 834 F.2d 1334, 1344-45 (7th Cir.1987) (declining to find misconduct where the government disbursed, without recovering, a 3.5-gram sample of cocaine while negotiating sales aggregating up to 35 kilos); United States v. Buishas, 791 F.2d 1310, 1314 (7th Cir.1986) (similar; government disbursed, without recovering, a 69-gram sample of marijuana in the course of closing an 89-kilo deal). Although Valona and Buishas are structurally analogous to the case at hand, the government concedes that the quantity of drugs given to Fuentes is, in absolute terms, unprecedented. The question, then, is whether, at some point, the quantity of drugs disbursed on the government’s behalf may *6 become so large that, given all the attendant circumstances, the government’s role becomes qualitatively different, i.e., outrageous. The court below devised a seven-part test and, applying that test, determined the government’s actions to be outrageous. See Santana, 808 F.Supp. at 81-86. The court focused on (1) the type of drug furnished; (2) the sample’s potency or purity; (3) its relative size; (4) whether the defendant requested it; (5) whether the drugs were recovered; (6) what likely happened to them; and (7) whether the sample itself constitutes the corpus delicti of the crime charged in the indictment. 8 Id. at 81-82. We appreciate the district court’s effort to structure the exercise of judicial discretion, and we realize that the court did not intend its compendium to be exhaustive. See id. at 82. Nevertheless, we do not think that the inquiry into outra-geousness can usefully be broken down into a series of discrete components. Almost by definition, the power of a court to control prosecutorial excesses through resort to substantive aspects of the due process clause is called into play only in idiosyncratic situations — and such situations are likely to be highly ramified. Where facts are critically important and fact patterns tend to be infinitely diverse, adjudication can often best proceed on a case-by-case basis. The outra-geousness defense falls into this category. Thus, it is unproductive to force the determination of outrageousness into a mechanical mode. Let us be perfectly plain. We find that outrageousness, by its nature, requires an ad hoc determination. We do not suggest, however, that the assessment should be wholly unguided. The calculus must be rooted in the record, and it will often be informed by the various factors enumerated in the district court’s test, the DEA’s test, see supra note 8, and similar tests produced by other sources. 9 At bottom, however, outra-geousness is a concept, not a constant. What shocks the conscience in a given situation may be acceptable, though perhaps grim or unpleasant, under a different set of circumstances. Slashing a person’s throat with a sharp knife may be an unrelievedly outrageous course of conduct if one thinks in terms of Jack the Ripper, helpless women, and the shadowy streets of London; the same behavior will be thoroughly acceptable, however, if the knife is a scalpel, the knife-wielder a skilled surgeon performing a tracheotomy, the target a patient, and the venue an operating room. Although we recognize that formulaic tests offer administrative convenience and ease in application, we also recognize that neither life nor law can always be made convenient and easy. So here: there is simply no way to reduce the myriad combinations of potentially relevant circumstances to a neat list of weighted factors without losing too much in the translation. Cf. Borden v. Paul Revere Life Ins. Co., 935 F.2d 370, 380 (1st Cir.1991) (discussing “out-rageousness” in the context of tort liability and concluding that “[tjhere is no universal litmus test that a court can utilize to determine whether behavior is extreme and outrageous”). In addition to relying on a tightly structured formulation in an area of the law demanding flexibility, the district court com *7 pounded its error by omitting from that formulation a salient set of considerations: it disregarded the nature and scope of the ongoing investigation. The outrageousness vel non of a police officer’s actions can only be evaluated by taking into account the totality of the relevant circumstances. When the officer is on the trail of a criminal enterprise, these circumstances include the identity of the suspects, the gravity of past crimes, and the dangers foreseeably attributable to the enterprise’s uninterrupted progress (including likely future crimes). In this instance, the government had a solid basis to believe that Santana’s network could distribute up to 200 kilograms of heroin per month. Seen in that light, it does not shock our collective conscience to think that a lawman would dangle 13.3 grams of heroin as bait to land such a large-scale ring, even though delivery of the sample ran a palpable risk of ushering it into the marketplace. The district judge refused to honor this argument, which the magistrate described as setting “a big hook to catch a big fish,” for several reasons. We find none of them convincing. First, the judge worried that the big hook/big fish approach would remove any outer limit on “the quantity of drugs that the government can introduce to society.” Santana, 808 F.Supp. at 83. It is a sufficient answer to this concern that, here, the size of the sample was proportionate both to the perceived threat posed by the ongoing criminal activity and to the exigencies of the chase. Other eases, involving greater quantities of drugs or materially different circumstances, need not be decided unless and until they arise. Second, the judge concluded that “the government’s conduct served only to increase the aggregate sum of heroin available for consumption.” Id. at 84. This statement, which we read as a bid to repudiate the magistrate’s implicit assessment of costs and benefits, is highly questionable. Let us compare two worlds. In the first world, the government distributes 13.3 grams of heroin, but Santana’s network is put out of business. In the second world, the government exercises greater restraint in its undercover activities, but fails to gather enough evidence to immobilize the ring. The aggregate supply of heroin will be greater in the first world only if one is prepared to indulge the unlikely assumption that some other equally skilled criminal network will instantaneously pick up the slack. Third, the judge, without saying so in haec verba, seemingly suggests that some situations cannot be analyzed in terms of societal costs and benefits. See id. at 85; cf. Richard C. Donnelly, Judicial Control of Informants, Spies, Stool Pigeons, and Agents Provocateurs, 60 Yale L.J. 1091, 1111 (1951) (denouncing “the sinister sophism that the end, when dealing with known criminals or the ‘criminal classes,’ justifies the employment of illegal means”). We do not share the district court’s discomfiture with means/ends rationality or — what amounts to the same thing— cost/benefit analysis. At least when the deci-sionmaker uses a common currency of exchange and operates under conditions of reasonable certainty, cost/benefit analysis is a perfectly legitimate mode of legal reasoning, frequently employed by both courts and agencies. See generally Richard A. Posner, The Problems of Jurisprudence 105-08 (1990). Using such an approach here does not strike us as either theoretically unsound or fundamentally unfair. More important still, we can identify no constitutional impediment to the government weighing the risk of an immediate 13.3-gram increase in the heroin supply against the potential benefit of diverting vast quantities of heroin from the American market. The district court’s resistance to cost/benefit analysis is carried to its logical conclusion by appellee Fuentes. He maintains that no possible prosecutorial objective can justify the distribution of so much heroin by the government. But, since there is abundant precedent for distribution of drugs by law enforcement agents mounting stings and other undercover operations, see eases cited supra p. 5, the only course of action compatible with Fuentes’s argument would be to construct a per se rule, drawing a bright line at some particular quantity of drugs and forbidding lawmen to cross that line in dealing with suspected drug traffickers. We regard a per se rule in this context as unprecedented, *8 unworkable, 10 unwise, and thoroughly uninviting. We, therefore, refuse to travel that road. Saying that we reject the district court’s objections to the big hook/big fish metaphor is not tantamount to saying that we unreservedly embrace the comparison. A hook, regardless of its size, causes injury only to the fish that is caught. We think that a more useful metaphor is that it takes a wide net to catch a big fish. Of course, a net cast to catch a big fish (thought to be predatory) might also catch hundreds of relatively innocent little fish. But, if the big fish would have devoured millions of little fish, even the most tender-hearted marine biologist would be hard pressed to argue against the fisherman’s use of the net. In the final analysis, probing the magistrate’s metaphor for impre-cisions does not assist appellees’ cause, but, rather, reinforces our conviction that the intuition underlying the metaphor is sound. We have trolled enough in these waters. We conclude that, on the facts of this case, the district court erred in discounting the import of the criminal enterprise’s scope and the magnitude of the threat that it posed. This error possesses decretory significance: once the size of the sample is measured in relative rather than absolute terms, the investigation reviewed here is no longer unprecedented and the conduct in question cannot plausibly be classified as outrageous. 11 B. Misconduct Not Injuring Defendants. Generally speaking, an outrageous misconduct defense can prosper only if a defendant’s due process rights have been violated. The defense is normally not available in situations where the government has injured only third parties or committed a victimless gaffe. We would be compelled to reverse the ruling below on this basis even if the government’s deportment failed the test of outrageousness. In an early entrapment case, Justice Bran-déis wrote: “The prosecution should be stopped, not because some right of th[e] defendant’s] has been denied, but in order to protect the Government. To preserve it from illegal conduct of its officers. To protect the purity of its courts.” Casey v. United States, 276 U.S. 413, 425, 48 S.Ct. 373, 376, 72 L.Ed. 632 (1928) (Brandeis, J., dissenting). The obvious implication of this perspective—with its emphasis on the rule of law rather than on individual rights—is that the state ought not profit by its miscreancy, regardless of whether a charged defendant has been wronged. Although the doctrinal view of entrapment based on this philosophy never prevailed, see Russell, 411 U.S. at 428-36, 93 S.Ct. at 1641-45, the Second Circuit subsequently flirted with the same perspective in a different context. In an outrageous misconduct case decided on other grounds, Judge Friendly expressed tentative support, in the abstract, for the view that the government ought not reap prosecutorial success growing out of the seeds of misconduct injuring third parties. See United States v. Archer, 486 F.2d 670, 676-77 (2d Cir.1973). 12 The court below believed this principle to be applicable here. See Santana, 808 F.Supp. at 84-85. We do not agree. *9 In our estimation, the Archer dictum is incompatible with later pronouncements of the Supreme Court. The flagship case is United States v. Payner, 447 U.S. 727, 100 S.Ct. 2439, 65 L.Ed.2d 468 (1980). There, the government obtained evidence against a defendant by rifling a third party’s briefcase. Although no due process claim was presented on appeal, the Court seized the occasion to address the precise question of misconduct injuring third parties and adopted a distinction first endorsed by the Hampton plurality: [E]ven if we assume that the unlawful briefcase search was so outrageous as to offend fundamental “ ‘canons of decency and fairness,’” Rochin v. California, 342 U.S. 165, 169 [72 S.Ct. 205, 208, 96 L.Ed. 183] (1952) ... the fact remains that “[t]he limitations of the Due Process Clause ... come into play only when the Government activity in question violates some protected right of the defendant.” Hampton v. United States, supra, [425 U.S.] at 490 [96 S.Ct. at 1650] (plurality opinion). Payner, 447 U.S. at 737 n. 9, 100 S.Ct. at 2447 n. 9 (1979). This statement, to be sure, is dictum—but it bears the earmarks of deliberative thought purposefully expressed. The statement is clear, pointed, and subscribed to by a 6-3 majority of the Justices. It is also prominent in its placement, appearing, as it does, in the concluding footnote of a major opinion. What is more, the issue that footnote 9 addressed had been thoroughly debated in the recent past, the Payner dissent treated it as purporting to establish a “standing” limitation, see id. at 749 n. 15, 100 S.Ct. at 2453 n. 15 (Marshall, J., dissenting), and the footnote’s message has not been diluted by any subsequent pronouncement. Carefully considered statements of the Supreme Court, even if technically dictum, must be accorded great weight and should be treated as authoritative when, as in this instance, badges of reliability abound. See McCoy v. Massachusetts Inst. of Technology, 950 F.2d 13, 19 (1st Cir.1991) (concluding that “federal appellate courts are bound by the Supreme Court’s considered dicta almost as firmly as by the Court’s outright holdings, particularly when ... a dictum is of recent vintage and not enfeebled by any subsequent statement”) (collecting cases to like effect from other circuits), cert. denied, — U.S. -, 112 S.Ct. 1939, 118 L.Ed.2d 545 (1992); see also Charles Alan Wright, The Law of the Federal Courts § 58, at 374 (4th ed. 1983). We need not decide whether Payner established a limitation on standing in the strict sense of the word, or merely signaled that defendants are highly unlikely to prevail when they seek to vindicate the rights of third parties. In either event, Payner makes manifest that, here, the trial court lacked authority under the due process clause to dismiss a charge on the basis that governmental misconduct caused conscience-shocking harm to non-defendants. See United States v. Valdovinos-Valdovinos, 743 F.2d 1436, 1437-38 (9th Cir.1984) (per curiam) (rejecting an outrageous misconduct defense on the strength of footnote 9 in a case in which government agents, trying to trap professional middlemen, lured illegal immigrants to the U.S. only to deport them), cert. denied, 469 U.S. 1114, 105 S.Ct. 799, 83 L.Ed.2d 791 (1985); United States v. Miceli, 774 F.Supp. 760, 770 (W.D.N.Y.1991) (rejecting an outrageous misconduct defense on the strength of footnote 9 in a case in which a government investigator seduced the defendant’s ex-wife in order to gather incriminating information about the defendant). IV. SUPERVISORY POWER The district court grounded its dismissal of count 3 on its supervisory power as well as on the due process clause. See Santana, 808 F.Supp. at 86. In a reprise of an argument earlier advanced, see supra Part III(B), the government asserts that a federal court’s supervisory power does not enable it to curb misconduct that injures only third parties by dismissing charges against uninjured defendants. We test this assertion. The contours of a court’s supervisory power are not much in doubt. Under them, a federal court “may, within limits, formulate procedural rules not specifically required by the Constitution or the Congress.” United States v. Hasting, 461 U.S. 499, 505, 103 S.Ct. 1974, 1978, 76 L.Ed.2d 96 (1983). The Hasting Court flagged three underlying purposes that can justify the use of supervisory *10 power in response to case-related misconduct, viz.: “to implement a remedy for violation of recognized rights; to preserve judicial integrity by ensuring that a conviction rests on appropriate considerations validly before the jury; and finally, as a remedy designed to deter illegal conduct.” Id. (citations omitted). While we have expressed the view that courts should be willing to “consider invoking [their] supervisory powers to secure enforcement of ‘better prosecutorial practice and reprimand of those who fail to observe it,’ ” United States v. Osorio, 929 F.2d 753, 763 (1st Cir.1991) (citation omitted), we have repeatedly cautioned that such powers must be used sparingly, see, e.g., id.; United States v. Babb, 807 F.2d 272, 279 (1st Cir.1986); United States v. Lieberman, 608 F.2d 889, 899 (1st Cir.1979), cert. denied, 444 U.S. 1019, 100 S.Ct. 673, 62 L.Ed.2d 649 (1980). Potent elixirs should not be casually dispensed. We do not believe that the circumstances of this case warrant such strong medicine. Although resort to a court’s supervisory power has not been foreclosed altogether as a means to remedy government misconduct not injuring the defendant, the Supreme Court has plainly semaphored its likely disapproval in several analogous contexts. For example, the Payner Court concluded that “the supervisory power does not authorize a federal court to suppress otherwise admissible evidence on the ground that it was seized unlawfully from a third party not before the court.” Payner, 447 U.S. at 735, 100 S.Ct. at 2446. In reaching this conclusion, the Court emphasized that such evidence could not be suppressed under the Fourth Amendment, see Rakas v. Illinois, 439 U.S. 128, 133-38, 99 S.Ct. 421, 425-28, 58 L.Ed.2d 387 (1978), and reasoned that the lower court’s choice of a different analytic framework did nothing to alter the relative values assigned to the underlying interests. Payner, 447 U.S. at 736, 100 S.Ct. at 2447. The lesson that this portion of Payner teaches is that, in a case-specific context, society’s interest in adjudi-eating guilt and innocence on full information outweighs its interest in punishing governmental misconduct directed against third parties. The Court subsequently held that the supervisory power could not be invoked to reverse a conviction in order to castigate the prosecution for misconduct that did not prejudice (as opposed to injure) the defendant. 13 See Hasting, 461 U.S. at 505, 103 S.Ct. at 1978. Because the prosecutor’s actions in Hasting constituted harmless error vis-a-vis the defendant, see id. at 507, 103 S.Ct. at 1979, no relief was warranted. The holding of Hasting replicates the message sent by Payner, but it does so a fortiori: if society’s interest in fully informed adjudication sometimes can outweigh its interest in protecting the Fifth Amendment rights of defendants, then surely it can outweigh society’s more generalized interest in making law enforcement officers toe the line. The reasoning of the Hasting Court is also instructive. As in Payner, the Court in Hasting reasoned that when courts exercise the supervisory power, they must respect the balance of interests struck by conventional application of the legal doctrines governing the particular problem in the particular case. See id. at 505, 103 S.Ct. at 1978. Furthermore, the Hasting Court identified three justifications, or goals, in service of which the supervisory power might appropriately be invoked, see id. at 506-07, 108 S.Ct. at 1979; see also supra p. 9, and rested its holding in part on an analysis of them. The Court concluded that none of these three goals are significantly advanced when the error that is alleged to constitute misconduct proves harmless, for concerns over individual rights and the integrity of the judicial process are less acute in all such cases. See id. at 506, 103 S.Ct. at 1979. The Court stated that the third doctrinal goal — the deterrence of misconduct 14 — “is an inappropriate basis for re *11 versal where ... the prosecutor’s remark is at most an attenuated violation of [defendant’s right to remain silent] and where means more narrowly tailored to deter objectionable prosecutorial conduct are available.” Id. Another ease delineating limits on the supervisory power is Bank of Nova Scotia v. United States, 487 U.S. 250, 108 S.Ct. 2369, 101 L.Ed.2d 228 (1988). There, the Court ruled that, “as a general matter, a district court may not dismiss an indictment for errors in grand jury proceedings unless such errors prejudiced the defendants.” Id. at 254, 108 S.Ct. at 2373. In reaching this conclusion, the Court adverted to PaynePs point that value choices dictated by the resolution of the underlying legal problem should not be affected by the source from which an inquiring court draws its power. See id. at 255, 108 S.Ct. at 2373-74. The Court also reaffirmed Hasting’s point that the rationales for invoking supervisory power are much weaker in the harmless error context. 15 See id. at 255-56, 108 S.Ct. at 2373-74. In keeping with the Supreme Court’s teachings, this court has repeatedly refused to sanction the deployment of supervisory power in order to redress harmless error. See Osorio, 929 F.2d at 763 (finding no nexus between the alleged misconduct and any prejudice to the defendant); United States v. Pacheco-Ortiz, 889 F.2d 301, 310 (1st Cir.1989) (denying relief when prejudice was not a “product” of alleged misconduct); Babb, 807 F.2d at 272; Lieberman, 608 F.2d at 899; see also United States v. Hastings, 847 F.2d 920, 927 (1st Cir.), cert. denied, 488 U.S. 925, 109 S.Ct. 308, 102 L.Ed.2d 327 (1988). We think this line of cases adequately evinces our institutional belief that, taken together, Payner, Hasting, and Bank of Nova Scotia form a trilogy admonishing federal courts to refrain from using the supervisory power to conform executive conduct to judicially preferred norms by dismissing charges, absent cognizable prejudice to a particular defendant. 16 Accord United States v. Williams, 874 F.2d 968, 976 n. 23 (5th Cir.1989). Here, appellees sustained no redressable injury attributable to governmental misconduct. Accordingly, the district court erred as a matter of law when it invoked supervisory power to dismiss count 3 of the indictment. Before departing from these shores, we pause to add a qualification: the use of supervisory power to dismiss an indictment, in the absence of injury to the defendant, may not be entirely a dead letter. The Court’s reasoning in Hasting may be read to leave open the possibility that the goal of deterring future misconduct would justify using the supervisory power to redress conduct not injuring defendants if the conduct is plainly improper, indisputably outrageous, and not redressable through the utilization of less drastic disciplinary tools. See Hasting, 461 U.S. at 506, 103 S.Ct. at 1979. Be that as it may, we leave the qualification’s fate and dimensions for another day, as this is plainly not such a case. V. CONCLUSION In summary, the orphan doctrine of outrageous misconduct finds no nurturing home on the facts of this case because the objects of the government’s ongoing investigation satisfactorily justified whatever harm stemmed from the delivery (and subsequent loss) of a large heroin sample, and because, in any event, that harm was not incurred by the *12 appellees themselves. In like manner, because the trial court overestimated the reach of its supervisory power in cases of misconduct not injuring defendants, its alternative rationale crumbles. If there are exceptions to the general rules that we have elucidated — a matter on which we do not opine— they are assuredly not triggered by this case. Hence, the court lacked a sufficient legal basis for dismissing count 3 of the indictment. We need go no further. Although the effect of our ruling is to uphold the government’s tactics in this case, we remain secure in the knowledge that, despite restrictions hobbling the outrageous misconduct doctrine, law enforcement practices are subject to a wide range of specific “constitutional and statutory limitations and to judicially fashioned rules to enforce those limitations.” Russell, 411 U.S. at 435, 93 S.Ct. at 1644; cf. Hasting, 461 U.S. at 506 n. 5, 103 S.Ct. at 1979 n. 5 (illustrating more narrowly tailored means to punish prosecutorial misconduct). Moreover, the outrageous misconduct doctrine, no matter how cramped its confines, is not entirely mummified. Should the occasion and the necessity arise, we continue to believe that the law will prove itself adequate to the task of preventing the government from going too far. In the war on crime, as in conventional warfare, some tactics simply cannot be tolerated by a civilized society. Reversed. Footnotes [1] The heroin sample comprises about 2,500 doses of the size and purity typically sold on the street. See Gerald F. Uelman & Victor G. Haddox, Drug Abuse and the Law Sourcebook, § 2.4(a) at 2-19 (1991). The DEA authorized delivery pursuant to section III-E of the DEA’s Domestic Operations Guidelines, 20 Crim.L.Rep. (BNA) 3055-58 (Feb. 2, 1977). [2] Count 3 named only Santana and Fuentes. Hence, they are the lone appellees. [3] The briefs highlight several other factual disputes. By and large, these disputes hinge on the admissibility of an affidavit appended to the motion to reconsider — an affidavit which tries to shed light on the sample's ultimate disposition and the agent's motive in delivering it. Because this affidavit was not proffered originally, and because the lower court made no findings concerning it, we consider only two undisputed portions of the affidavit, namely, that the agent, in asking his superiors to arrange for a sample, believed that "Fuentes was testing whether I was a real drug dealer,” and that supplying the sample "was an important part of Fuentes’ evaluation whether to go forward with the deal.” We will assume, as appellees implore, that most, if not all, of the 13.3 grams of heroin reached end users. [4] In Hampton, a concurrence combined with the plurality to reject the appeal. However, the two concurring Justices switched sides to form a different majority vivifying the doctrine of outrageous misconduct. See Hampton, 425 U.S. at 491-95, 96 S.Ct. at 1650-53 (Powell, J. concurring). [5] In Judge Easterbrook’s view, the appropriateness of the government’s decision to supply drugs as part of an undercover operation presents a "political” question that is quintessentially nonjusticiable. Miller, 891 F.2d at 1272. With respect, we think this conceptualization stretches the military analogy too far. We adhere instead to the idea that "those charged with th[e] investigative and prosecutorial duty should not be the sole judges of when to utilize constitutionally sensitive means in pursuing their tasks.” United States v. United States District Court, 407 U.S. 297, 317, 92 S.Ct. 2125, 2136, 32 L.Ed.2d 752 (1972). [6] In addition to Twigg, one court of appeals invoked the doctrine in an alternative holding, see United States v. Lard, 734 F.2d 1290, 1296 (8th Cir.1984), and another directed the district court to determine whether outrageous misconduct should be found on remand, see Bogart, 783 F.2d at 1438. A smattering of district courts have also applied the outrageous misconduct doctrine to the defendant’s advantage. See, e.g., United States v. Marshank, 777 F.Supp. 1507, 1524 (N.D.Cal.1991); United States v. Gardner, 658 F.Supp. 1573, 1577 (W.D.Pa.1987); United States v. Batres-Santolino, 521 F.Supp. 744, 751-52 (N.D.Cal.1981). [7] We recognize that narcotics differ in many ways, including size, weight, and potency; and that, therefore, a small amount of a particular drug, say, heroin, may be much more lethal than a larger amount of a different drug, say, marijuana. For ease in reference, however, we use the term “quantity” throughout this opinion as a proxy for dangerousness. [8] In contrast, the relevant DEA guidelines, see supra note 1, suggest consideration of (1) the type and amount of the drug contained in the sample; (2) the likelihood that the sample will reach consumers; (3) the number and prominence of the suspects implicated; (4) the type and amount of evidence needed to complete the ongoing investigation; (5) the time required to do so; and (6) the likelihood of obtaining such evidence. Although the DEA's list, like the district court’s list, contains factors relevant to the seriousness of harm likely to be suffered by end users, the DEA’s list emphasizes, and the court's list slights, the likelihood that the investigation will lead to the prosecution of important drug dealers. [9] See, e.g., United States v. Feinman, 930 F.2d 495, 498 (6th Cir.1991) (suggesting that a reviewing tribunal weigh (1) the importance of the investigation, evidenced by the type of criminal activity targeted, (2) whether the criminal enterprise predated the investigation, (3) whether the investigator directed or controlled the criminal activity, and (4) the investigation's impact on the commission of the crime); United States v. Gardner, 658 F.Supp. 1573, 1576-77 (W.D.Pa.1987) (suggesting that a reviewing tribunal weigh (1) the government's role in creating the crime, (2) the illegality or immorality of the police conduct, (3) the defendant's predisposition to commit the crime, and (4) whether the investigation is aimed at preventing further criminality). [10] We illustrate one of the many problems that such a per se rule would present. Were we to draw such a line at, say, 10 grams of heroin, we would be handing criminals a foolproof way to detect whether prospective new suppliers were actually government agents: simply demand a sample equal to 11 or 12 grams of heroin. [11] We do not totally reject the possibility, suggested by the court below, that outrageous misconduct may be found apart from situations in which the government has used brutality or induced commission of a crime. We simply note that the case at hand does not require us to explore this doctrinal frontier. [12] Two recent Second Circuit cases cite Archer in connection with the proposition that courts "will closely examine those cases in which the Government misconduct injures third parties in some way.” United States v. Thoma, 726 F.2d 1191, 1199 (7th Cir.), cert. denied, 467 U.S. 1228, 104 S.Ct. 2683, 81 L.Ed.2d 878 (1984); accord United States v. Chin, 934 F.2d 393, 400 (2d Cir.1991). But neither panel actually applied this principle, because no injury to third parties had been established. By like token, in United States v. Panet-Collazo, 960 F.2d 256 (1st Cir.), cert. denied, — U.S. -, 113 S.Ct. 220, 121 L.Ed.2d 158 (1992), we were able to sidestep the issue because the heroin sample provided by the government as part of the sting was not used in a manner outrageously injurious to third parties. See id. at 260. [13] Misconduct not injuring the defendant is a subset of harmless error (which itself might be described as misconduct not prejudicing the defendant). For our purposes, the two categories may be fruitfully analyzed as one. The only salient difference between them is that the larger set subsumes not only misconduct that injures third persons and victimless misconduct, but also subsumes misconduct that violates a defendant's rights without affecting the outcome of his case. [14] We highlight this goal because it not only constitutes the linchpin of the district court's rationale for employing the supervisory power in *11 this case, but also serves as the mainstay of the supporting arguments advanced by the appellees and by the amici. [15] It is a short step, sideways rather than forward, from Hasting to Bank of Nova Scotia. Hasting holds that the supervisory power may not be used to evade the constitutional harmless error doctrine of Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967); Bank of Nova Scotia holds that the supervisory power may not be used to evade the less searching harmless error inquiry mandated by Fed. R.Crim.P. 52(a). [16] The Second Circuit has gone even further, reading the Supreme Court's cases to suggest that "the federal judiciary’s supervisory powers over prosecutorial activities that take place outside the courthouse is extremely limited, if it exists at all.” United States v. Lau Tung Lam, 714 F.2d 209, 210 (2d Cir.), cert. denied, 464 U.S. 942, 104 S.Ct. 359, 78 L.Ed.2d 322 (1983). Because the case at bar does not require that we probe the ramifications of this suggestion, we take no view of it.