Before Boudin, Chief Judge, Selya, Senior Circuit Judge, and Gelp¡,1 District Judge.
A claim of selective prosecution depends in large part on a defendant's ability to prove that the government has treated him differently from similarly situated offenders. The appeal before us turns on what factors courts should take into account in configuring the pool of similarly situated offenders for purposes of this comparison. The tale follows.
In the underlying case, the government charged defendant- appellant Samuel J. Lewis, also known as Shaheed Lewis, with making false statements on multiple federal firearms applications. The defendant moved to dismiss the indictment on selective prosecution grounds, asserting that the government had elected to prosecute him but not others who were similarly situated because he is African-American and Muslim. Relatedly, he moved for discovery in aid of his selective prosecution theory and for an evidentiary hearing. The district court denied all of these motions.
Following his conviction on various counts, the defendant appeals from the denial of his pretrial motions. The issue on appeal boils down to whether discovery should have been permitted - an issue that hinges largely on whether the defendant mustered some evidence that the government had eschewed prosecution of similarly situated offenders. Discerning no error in the trial court's configuration of the pool of similarly situated offenders and no abuse of discretion in its refusal to allow discovery, we affirm the judgment below.
The facts are not particularly complicated. The defendant is an African-American Muslim man who, up until the events in question here, had an unblemished record. Over a three-year span from November of 2000 to October of 2003, the situation changed.
During that period, the defendant obtained approximately thirty-two firearms. From August of 2002 through September of 2003, he procured no fewer than fifteen of those guns by providing - or so the government alleged - false residence addresses on a government form (ATF Form 4473) comprising part of the standard federal firearms application.2 In two additional instances he allegedly acted as a straw purchaser, buying a gun for a friend but representing on the application that he was acting for himself. Knowingly making a misrepresentation in the procurement of a firearm constitutes a federal felony. See 18 U.S.C. §§ 922(a)(6), 924(a)(1)(A).
The frequency of the defendant's arms purchases raised a red flag with the Bureau of Alcohol, Tobacco, Firearms, and Explosives (ATF). That federal agency, in concert with the inter-agency Joint Terrorism Task Force, launched an investigation. The investigators discovered obvious discrepancies relating to the residence addresses listed by the defendant in the various applications. They also learned that the defendant reportedly had changed his first name to Shaheed (which in Arabic means, among other things, "martyr"), that he had stated that he wished to move to a country in which there was a war or a crisis, and that he had mused about wanting to die in a jihad. One of the witnesses whom the investigators interviewed related that, in the course of firing a high-velocity weapon at a shooting range, the defendant had threatened him.
The witnesses also revealed that the defendant had visited either Syria or Somalia for approximately one month in 2003. Following up on this lead, the investigators unearthed independent evidence that the defendant had reported his passport missing. This is reputedly a familiar stratagem used to obtain a "clean" passport after visiting countries associated with terrorism.
Like other federal agencies, the ATF issues procedures to guide its agents in their investigation of criminal activity. The ATF's guidelines caution agents to evaluate informant credibility through "[a]ll reasonable efforts" before "committing substantial resources or taking significant enforcement action" in reliance on information from an informant "of unknown or untested reliability." Treas. Dep't, ATF Order 3210.7B, at 17 (June 28, 1989). The three principal witnesses interviewed by the investigators included the defendant's ex-wife, his current wife's ex-husband (who supposedly learned relevant information through conversations with a thirteen-year-old child), and an employee of a shooting range frequented by the defendant.
The probe culminated in a multi-count indictment. The bill, which originally contained twenty-one counts, was eventually winnowed to seventeen. Fifteen of these counts charged the defendant with making false statements about his place of residence on federal firearms applications in violation of 18 U.S.C. § 924(a)(1)(A). The remaining two counts charged him with illegally acting as a straw purchaser in violation of 18 U.S.C. § 922 (a)(6).
In due season, the defendant moved (i) to dismiss the charges on the ground of selective prosecution, (ii) to allow discovery in aid of the foregoing motion, and (iii) for an evidentiary hearing. The government opposed the motions and, as part of its opposition, submitted the affidavit of an ATF agent who described the results of the investigation in some detail. The district court denied the motions, concluding that the defendant had failed to present sufficient evidence of disparate treatment to warrant further inquiry. United States v. Lewis, Crim. No. 05-40001 (D. Mass. May 11, 2006) [2006 WL 4385752, at *7].
After a trial, a petit jury convicted the defendant on the fifteen "false statement" counts. At the same time, the jury acquitted him on the two "straw purchaser" counts. The district court sentenced the defendant to fifteen months in prison and three years of supervised release. This timely appeal ensued.
This is a rifle-shot appeal: the defendant candidly states that "[t]he district court's denial of defendant's request for further discovery in aid of his selective prosecution claim is the subject of this appeal." Appellant's Br. at 2. Moreover, this issue is the only one to which he devotes any developed argumentation. Consequently, his appeal stands or falls on the supportability vel non of the lower court's denial of the discovery motion.3
Our cases have used an abuse of discretion standard for appellate review of a claim that a trial court erred in refusing to allow discovery in aid of a selective prosecution defense. See, e.g., United States v. Magana, 127 F.3d 1, 9 (1st Cir. 1997); United States v. Penagaricano-Soler, 911 F.2d 833, 838 (1st Cir. 1990). Subject only to a few narrow exceptions (none of which is applicable here), the "law of the circuit" rule binds a court within a particular circuit to follow circuit precedent directly or closely on point. See United States v. Guzm n, 419 F.3d 27, 31 (1st Cir. 2005); United States v. Chhien, 266 F.3d 1, 11 (1st Cir. 2001). That rule would seem to dictate that an abuse of discretion standard should be employed in reviewing the denial of the defendant's discovery motion.
In seeming defiance of the law of the circuit rule, the defendant cajoles us to ignore circuit precedent in this instance and instead undertake de novo review. His cajolery has two ostensible justifications. First, he sings a siren song to the effect that we are not bound by our past opinions because we have not "explicitly" adopted the abuse of discretion standard in any of them. Second, he suggests that the better standard of review is not abuse of discretion but, rather, the de novo review standard employed by two of our sister circuits in comparable cases. See, e.g., United States v. James, 257 F.3d 1173, 1178 (10th Cir. 2001); United States v. Olvis, 97 F.3d 739, 743 (4th Cir. 1996).4 We find neither of these blandishments appealing.
The defendant's first initiative holds no weight. The fact that our earlier cases have not formally adopted a standard of review does not excuse us from following the law of the circuit rule. That rule is rooted in the need for consistency within a circuit; its force does not depend on a prior panel's use of talismanic phrases. So long as a prior panel, in a holding directly or closely on point, makes clear its choice of a rule of law, that choice is binding on newly constituted panels within the circuit, subject only to the isthmian exceptions noted in our earlier decisions.
The defendant's second initiative is even less attractive. The law of the circuit rule does not depend on whether courts outside the circuit march in absolute lockstep with in-circuit precedent. Thus, the emergence of a minority view, without more, does not open an escape hatch that justifies a newly constituted panel in repudiating an unbroken skein of circuit precedents. See, e.g., United States v. Rodriguez, 311 F.3d 435, 438-39 (1st Cir. 2002). We therefore reject the defendant's importunings and review the denial of his discovery motion for abuse of discretion.
The defendant, ably represented, has a fallback position related to the standard of review. He argues that, in applying an abuse of discretion standard, we ought to parse out the district court's determination regarding what constitutes a similarly situated pool of offenders and review that determination de novo. This argument has some theoretical underpinnings.
As we have explained in other cases, the abuse of discretion standard is not monolithic. See Roger Edwards, LLC v. Fiddes & Son Ltd., 427 F.3d 129, 132 (1st Cir. 2005); United States v. McIntosh, 380 F.3d 548, 553-54 (1st Cir. 2004). In practice, that standard contemplates de novo review of abstract questions of law,5 clear error review of findings of fact, and deferential review of judgment calls. Roger Edwards, 427 F.3d at 132.
This nuanced formulation of the standard of review does not help the defendant. The fact-intensive inquiry involved in determining who constitutes a similarly situated individual belies the defendant's assertion that the issue is invariably law-dominated. See United States v. Armstrong, 517 U.S. 456, 466 (1996) (stating that all facts must be considered to determine who is similarly situated); cf. Cordi-Allen v. Conlon, 494 F.3d 245, 251 (1st Cir. 2007) (explicating same proposition in civil "class of one" equal protection case). Here, moreover, all roads lead to Rome; as we explain below, the configuration of the class is sufficiently clear-cut that the precise standard of review makes no difference.
With these preliminary skirmishings behind us, we grapple with the merits of the defendant's assertion that he made a sufficient showing to justify discovery in aid of his selective prosecution theory. We approach this assertion mindful that federal prosecutors must be afforded substantial discretion not only in determining whether to prosecute a suspected violation of federal law but also in deciding what charges should be lodged. See Armstrong, 517 U.S. at 464; Wayte v. United States, 470 U.S. 598, 607 (1985). Once made, these decisions enjoy a presumption of regularity (which includes a presumption of good faith). See Armstrong, 517 U.S. at 464; United States v. Graham, 146 F.3d 6, 9 (1st Cir. 1998).
That presumption of regularity should not lightly be discarded for it serves important policy interests. In particular, the presumption enhances the efficacy of prosecutorial strivings to enforce the law while at the same time limiting courts' abilities to circumscribe executive authority in areas outside the realm of judicial competence. See Wayte, 470 U.S. at 607-08; United States v. Bourgeois, 964 F.2d 935, 939 (9th Cir. 1992). It is, therefore, unsurprising that the presumption is formidable; it can be overcome only by a proffer of "clear evidence" that the prosecutor acted impermissibly in pursuing a case. Armstrong, 517 U.S. at 465.
A showing of selective prosecution can, of course, undercut the presumption of regularity. The essence of such a showing is that a prosecutor has pursued a case for a constitutionally impermissible reason, such as the defendant's race, religion, or other characteristic cognizable under equal protection principles. Wayte, 470 U.S. at 608. Carrying this burden entails a binary showing: the defendant must adduce clear evidence of both the discriminatory effect of the prosecution and the prosecutor's discriminatory intent. Armstrong, 517 U.S. at 465.
The evidentiary threshold that a defendant must cross in order to obtain discovery in aid of a selective prosecution claim is somewhat below "clear evidence," but it is nonetheless fairly high. Id. at 468. To cross this lower threshold, a defendant must present "some evidence" tending to show both discriminatory effect and discriminatory intent. Id. (citing United States v. Berrios, 501 F.2d 1207, 1211 (2d Cir. 1974)). It follows that discovery will not be allowed unless the defendant's evidence supports each of the two furcula of his selective prosecution theory: failure on one branch dooms the discovery motion as a whole. See, e.g., United States v. Bass, 536 U.S. 862, 863-64 (2002) (per curiam).
"Some evidence" is admittedly a protean standard. For this purpose, the evidence in support of the asserted discriminatory effect must comprise a credible showing that similarly situated individuals who do not share the protected characteristic were not prosecuted. Armstrong, 517 U.S. at 469. Similarly, the evidence in support of the asserted discriminatory intent must consist of a credible showing that the government chose to prosecute "at least in part because of, not merely in spite of," the defendant's protected characteristic. Wayte, 470 U.S. at 610 (quoting Personnel Adm'r of Mass. v. Feeney, 442 U.S. 256, 258 (1979)).
In the case at hand, the defendant claims selective prosecution based on an amalgam of race and religion - he is an African- American Muslim - and alleges that his prosecution had a discriminatory effect and was spurred by discriminatory intent.
The district court disagreed. In denying his discovery motion, the court applied the standards alluded to above. It determined that the pool of offenders situated similarly to the defendant consisted of non-African-Americans and/or non-Muslims who had committed multiple misrepresentation offenses in connection with firearms paperwork, who posed a danger of violence, and who may have had links to terrorism. Lewis, 2006 WL 4385752, at *7. After canvassing the evidence that had been submitted, the court found that the defendant had misconceived the dimensions of the pool of similarly situated offenders and that, by failing to tender any proof that non-African-Americans and/or non- Muslims in the appropriate pool had been spared from prosecution, the defendant had failed to cross the discovery threshold. Id. Finally, despite the government's reliance on the three witnesses whose credibility the defendant questioned, the court found that the evidence was not so flimsy as to constitute a showing of discriminatory intent. Id.
The defendant mounts two primary challenges to the district court's decision. First, he calumnizes the court's configuration of the pool of similarly situated offenders. Second, he faults the court for neglecting to find that he had made a sufficient showing of discriminatory intent. We approach these challenges one by one.
We start with discriminatory effect. The defendant insists that the district court framed the contours of the pool of similarly situated offenders too narrowly and, based on this cramped configuration, erroneously concluded that the defendant had not provided any evidence of discriminatory effect. Specifically he asserts that the appropriate pool should be composed of non-African-American, non-Muslim persons who have submitted inaccurate paperwork in connection with federal firearms applications. He mentions that by considering two additional criteria - multiple violations and potential terrorist connections - the court incorrectly circumscribed the pool of similarly situated offenders and, in the process, set the bar for discovery too high.
Refined to bare essence, the defendant's thesis is that the pool of similarly situated offenders should be populated by those who have committed the same basic offense - no more and no less. On the assumption that this approach was correct, he introduced a series of analyses showing that over a three-year period no one else had been prosecuted in the District of Massachusetts for a weapons-related offense as picayune as misstating an address on a federal firearms application (at least when neither the purchaser of the gun nor its intended user had a prior criminal record). Building on this foundation, he argues that statistical probability indicates that while white non-Muslims quite probably have committed such minor infractions, they have not been prosecuted. And, finally, he asserts that in any event his prosecution is inconsistent with the normal practices followed by the U.S. Attorney's Office in the District of Massachusetts - practices that emphasize prosecution of crimes with different characteristics.
The government's response is terse and to the point. It says that a broader array of circumstances must be factored into the mix. Once that is done, the defendant's prosecution may seem unique, but the sheer number of misrepresentation offenses and the concomitant evidence of terrorism links distinguish this case and make prosecution permissible. In other words, it insists that the district court properly considered these distinguishing factors in configuring the pool of similarly situated offenders. Finally, it asserts that the defendant's evidence of prosecutorial priorities shows only what crimes had been prosecuted in the past, not some hard-and-fast institutional rule about which crimes should be prosecuted.
Prosecutorial decisions are often highly ramified, and courts must be chary of relying on raw statistics as purported proof of selective prosecutions. See Bass, 536 U.S. at 864; see also United States v. Thorpe, 471 F.3d 652, 658 (6th Cir. 2006) (applying this principle); United States v. Hedaithy, 392 F.3d 580, 607-08 (3d Cir. 2004) (same). Before reaching any questions related to the defendant's statistical evidence, however, we must elucidate the proper interpretation and application of the term "similarly situated." Only then can we judge the relevancy of the statistical proffer.
Although we have not previously provided a distinct definition of the term "similarly situated" in the selective prosecution context, classic equal protection principles light our path and limn the attributes of one who is similarly situated. See Armstrong, 517 U.S. at 465. A similarly situated offender is one outside the protected class who has committed roughly the same crime under roughly the same circumstances but against whom the law has not been enforced. See id. at 469. In configuring the pool of similarly situated offenders, "no fact should be omitted to make it out completely." Id. at 466 (quoting Ah Sin v. Wittman, 198 U.S. 500, 508 (1905)) (emphasis in original).
To be sure, this statement cannot be taken literally. The focus of an inquiring court must be on factors that are at least arguably material to the decision as to whether or not to prosecute. Material prosecutorial factors are those that are relevant - that is, that have some meaningful relationship either to the charges at issue or to the accused - and that might be considered by a reasonable prosecutor. Cf. Cordi-Allen, 494 F.3d at 250-51 (requiring comparators to be similarly situated "in all relevant aspects" in a civil "class of one" context); Perkins v. Brigham & Women's Hosp., 78 F.3d 747, 751 (1st Cir. 1996) (stating in employment discrimination context that comparator must be similarly situated in "material" respects). Unrelated, irrelevant, or trivial factors cannot meet the materiality requirement and, therefore, cannot be built into the configuration of the pool.
The bottom line, then, is that a district court should assess every material fact in rendering its judgment as to which offenders should be deemed similarly situated. See Olvis, 97 F.3d at 744 (explaining that "defendants are similarly situated when their circumstances present no distinguishable legitimate prosecutorial factors that might justify making different prosecutorial decisions with respect to them").
A multiplicity of factors legitimately may influence the government's decision to prosecute one individual but not another. See Wayte, 470 U.S. at 607; see also Magana, 127 F.3d at 9 (listing representative factors). These may include, inter alia, the comparability of the crimes, the similarities in the manner in which the crimes were committed, the relative efficacy of each prosecution as a deterrent, and the equivalency of the evidence against each prospective defendant. See United States v. Smith, 231 F.3d 800, 810 (11th Cir. 2000); see also United States v. Peterson, 233 F.3d 101, 105 (1st Cir. 2000) (considering individuals' relative levels of participation in illegal scheme).
In this case, the district court took account of these precepts and configured the pool of similarly situated offenders with reference to the nature and numerosity of the offenses and the incidence of possible links to terrorism. While the defendant labors to persuade us that this configuration is too specific, we are not convinced. Each of the items that the district court factored into the configuration calculus is relevant and material. Those criteria are, therefore, appropriate.
In reaching this conclusion, we do not write on a pristine page. Other courts have affirmed that prosecutors may permissibly consider specific factors such as those at issue here in deciding whom to prosecute. The Eleventh Circuit, for instance, has acknowledged that the government may lawfully single out an offender for prosecution because he has violated a law over and over again. See Smith, 231 F.3d at 812 (concluding that a prosecutor may "legitimately place a higher priority on prosecuting someone who commits an offense three, six or seven times, than someone who commits [that] offense once or twice, especially when the offense is a non-violent one").
By like token, courts have upheld the government's decision to prosecute more readily when the specter of terrorism is implicated. See, e.g., United States v. Khan, 461 F.3d 477, 498 (4th Cir. 2006) (affirming denial of discovery in selective prosecution case); cf. Reno v. Am.-Arab Anti-Discrim. Comm., 525 U.S. 471, 491-92 (1999) (holding that consideration of terrorist links in deportation context is not sufficient to constitute selective deportation); Kandamar v. Gonzales, 464 F.3d 65, 74 (1st Cir. 2006) (similar).
That ends this aspect of the matter. Here, the record reveals with conspicuous clarity that the defendant committed multiple violations of the firearms laws and that he did so in the shadow of evidence raising plausible concerns about his possible terrorist connections. The prosecutor had statements from at least three witnesses, together with independent evidence, to buttress these suspicions.6 On this substantial record, there is no principled way that we can find an abuse of discretion in the district court's configuration of the pool of similarly situated offenders.
Once we have determined that the district court properly configured the pool of similarly situated offenders, the rest is child's play. The record reveals no indication that any similarly situated white or non-Muslim person escaped prosecution. By like token, the record does not reveal that the prosecutors acted contrary to U.S. Attorney's guidelines or practices anent the prosecution of similarly situated offenders. Consequently, the trial court's holding that the defendant failed to present some evidence capable of demonstrating discriminatory effect is unassailable.
This brings us to the defendant's second argument: that the district court erred in accepting the credibility of the three witnesses whose statements influenced the government's decision to prosecute. Because the government relied on those witnesses in derogation of its standard practice, this thesis runs, its use of the information gleaned from them must have been a front for discriminatory intent.
The government counters that in this instance multiple sources provided similar types of information. It says that, taking account of this cross-corroboration, it appropriately concluded that the witnesses possessed sufficient credibility. Thus, the government argues, the district court committed no error in concluding that the prosecutors' actions did not evince any discriminatory intent.
We need not linger long over these conflicting contentions. The witness credibility arguments go to the second prong of the discovery inquiry: the presence or absence of discriminatory intent. Here, however, the defendant has failed to make the threshold showing required under the first prong of the test; he has failed to adduce some evidence that the prosecution manifested a discriminatory effect. For that reason, his appeal of the denial of the discovery motion founders, and further consideration of the evidence pertaining to discriminatory intent would serve no useful purpose. See Bass, 536 U.S. at 863.
We live in an era in which the incidence of violent crime is high and terrorism is a persistent threat. In that climate, a false statement in a firearms application raises a modicum of concern, and the level of concern escalates almost exponentially as the number of weapons and the number of false statements grow. When information about multiple guns and multiple false statements is enmeshed with information suggesting terrorist links, prosecutors can scarcely be criticized for moving forward. This prosecution may be unique - but that is because the compendium of material facts on which it rests is unique.
We need go no further. For the reasons elucidated above, we conclude without serious question that the district court did not abuse its discretion in denying the discovery motion. On this record, no credible claim of selective prosecution could lie.
1. Of the District of Puerto Rico, sitting by designation.
2. Each of these purchases was within the purview of the Gun Control Act of 1968, 18 U.S.C. §§ 921-930. That statute requires gun dealers to keep records of the information collected by means of ATF Form 4473.
3. This makes perfect sense: the quantum of evidence that would be needed to authorize discovery is less than the quantum of evidence needed to dismiss the indictment on selective prosecution grounds. See United States v. Armstrong, 517 U.S. 456, 465, 468 (1996). By the same token, if the defendant's initial proffer is insufficient to justify a discovery order, there would be no need for an evidentiary hearing.
4. This aspect of the defendant's argument ignores the fact that no fewer than five other circuits have recently endorsed abuse of discretion review in this context. See United States v. Thorpe, 471 F.3d 652, 657 (6th Cir. 2006) (collecting cases).
5. The understanding, of course, is that a mistake of law is always an abuse of discretion. See, e.g., Rosario-Urdaz v. Rivera- Hern ndez, 350 F.3d 219, 221 (1st Cir. 2003); United States v. Snyder, 136 F.3d 65, 67 (1st Cir. 1998).
6. We reject out of hand the defendant's effort to incorporate witnesses' credibility as part of his argument as to why potential terrorism links should not be included in the definition of similarly situated offenders. That position conflates discriminatory effect and discriminatory intent and, thus, confuses the two prongs of the discovery test.