United States Court of Appeals, Fourth Circuit.
*210 ARGUED: Joseph Bart Gilbert, Office of the Federal Public Defender, Raleigh, North Carolina, for David A. Passaro. Banumathi Rangarajan, Office of the United States Attorney, Raleigh, North Carolina, for the United States. ON BRIEF: Thomas P. McNamara, Federal Public Defender, James E. Todd, Jr., Research and Writing Specialist, Eric J. Brignac, Research and Writing Specialist, Office of the Federal Public Defender, Raleigh, North Carolina, for David A. Passaro. George E.B. Holding, United States Attorney, Anne M. Hayes, Assistant United States Attorney, Office of the United States Attorney, Raleigh, North Carolina, for the United States.
Before WILKINSON, MOTZ, and GREGORY, Circuit Judges.
Affirmed in part and vacated and remanded in part by published opinion. Judge MOTZ wrote the opinion, in which Judge WILKINSON and Judge GREGORY joined.
DIANA GRIBBON MOTZ, Circuit Judge:
This case arises from the conviction in a United States federal court of an American citizen for the brutal assault on an Afghan national in Afghanistan. A jury in the *211 Eastern District of North Carolina found David A. Passaro, a Central Intelligence Agency civilian contractor, guilty of assault on Abdul Wali. The assault occurred in 2003 at Asadabad Firebase, a United States Army outpost in Afghanistan.
Passaro asserts that American courts lack subject matter jurisdiction over assaults in Asadabad. He further maintains that his prosecution offends separation-of-powers principles and arises from statutes unconstitutionally vague as applied to him. Finally, Passaro challenges certain evidentiary, jury instruction, and sentencing rulings of the district court.
This case presents novel questions concerning the reach of federal criminal law to acts that an American civilian commits abroad while in service to this Country. After careful consideration, we reject all of Passaro's challenges to his conviction. We conclude, however, as Passaro and the Government both argue, that the district court erred in sentencing him. Accordingly, we affirm Passaro's conviction but vacate his sentence and remand for resentencing.
After the September 11, 2001, terrorist attacks, the United States conducted a military operation in Afghanistan in an effort to topple the Taliban regime. Sometime in late 2001, as part of this effort, American and coalition military troops forcibly obtained control of the Asadabad Firebase in north-east Afghanistan.
A thick, ten-foot-high mud wall surrounds Asadabad, which is an old fortress that covers approximately 25 acres. By May 2003, when Passaro arrived at Asadabad, the fortress contained approximately a dozen useable buildings, which coalition forces employed as offices, living quarters, and detention facilities. In addition, the United States Army had installed electricity and was in the process of providing fresh well water in Asadabad.
Coalition forces conducted military and intelligence operations from Asadabad and used the base to train Afghan militia personnel. At any one time, the United States stationed about 200 military personnel at the firebase, along with a number of paramilitary civilian contractors.
Passaro, a former Army special forces medic, arrived at Asadabad in May 2003 as part of this contingent of paramilitary contractors. At about the same time, military commanders had become concerned with rocket attacks on Asadabad, which occurred on a regular basis. The United States began to suspect that a local Afghan named Abdul Wali orchestrated the rocket attacks, and the military formulated a plan to capture Wali for questioning.
Before American forces could execute this plan, however, Wali voluntarily surrendered himself for questioning on June 18, 2003. After some initial interviews, American commanders at Asadabad decided to detain Wali. They imprisoned him in a detention cell, shackling his legs, binding his wrists together, and placing a hood over his head. The military placed him under 24-hour, two-person, armed guard.
Sometime on the evening of the next day, June 19, the CIA commander at Asadabad authorized Passaro to interrogate Wali. It is undisputed that for the next two days, Passaro "interrogated" Wali. This "interrogation" involved Passaro's brutal attacks on Wali, which included repeatedly throwing Wali to the ground, striking him open handed, hitting him on the arms and legs with a heavy, Maglite-type flashlight measuring over a foot long, and, while wearing combat boots, kicking Wali in the groin with enough force to lift him off the ground.
*212 Passaro "interrogated" Wali in this manner throughout the next day, June 20. And, although Wali's condition greatly deteriorated, Passaro continued the "interrogation" through the night of June 20. By June 21, Wali had lapsed into delirium, to the point where he twice asked his guards to shoot him and even lunged at a guard as if to take his gun. Later that day, while still in United States custody at Asadabad, Wali collapsed and died.
The next month, Passaro returned to North Carolina. A year later, a federal grand jury indicted him on two counts of assault with a dangerous weapon with intent to do bodily harm in violation of 18 U.S.C. § 113(a)(3) (2006) and two counts of assault resulting in serious bodily injury in violation of 18 U.S.C. § 113(a)(6) (2006).[1]
At trial, the translator present at the "interrogations" and the personnel charged with guarding Wali during them described in detail Passaro's brutal "interrogation" methods. Numerous witnesses, including Passaro's CIA supervisors, testified that no CIA official had encouraged or authorized Passaro's "interrogation" methods.
After an eight-day trial, the jury convicted Passaro of one count of felony assault resulting in serious bodily injury and three counts of the lesser-included offense of misdemeanor simple assault. The district court sentenced Passaro to a 100-month term of imprisonment, applying a number of sentencing enhancements and an upward departure to reach that result. Passaro noted a timely appeal challenging both the convictions and the sentence; the Government filed a cross-appeal as to the sentence.
Most fundamentally, Passaro challenges our subject matter jurisdiction over the crimes alleged against him. The Government predicated federal criminal jurisdiction in this case on the special maritime and territorial jurisdiction statute. See 18 U.S.C. § 7 (2006). This statute extends federal criminal jurisdiction to crimes, like assault, that states traditionally regulate, when the crimes occur in a federal enclave for example, United States military bases, federal buildings, and national parks, see id. § 7(3), the high seas and waters within this Country's admiralty and maritime jurisdiction, id. § 7(1), and certain aircraft and spacecraft, id. § 7(5)-(6). See United States v. Anderson, 391 F.3d 1083, 1086 (9th Cir.2004); United States v. Erdos, 474 F.2d 157, 159-60 (4th Cir.1973). In other words, the statute makes the site of the offense an element of the crime. See United States v. Brisk, 171 F.3d 514, 520 n. 4 (7th Cir.1999).
In 2001, Congress added an additional subsection to the special maritime and territorial jurisdiction statute. See Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism (USA PATRIOT) Act of 2001, Pub.L. No. 107-56, § 804, 115 Stat. 272, 377 (codified at 18 U.S.C. § 7(9) (2006)). This provision explicitly extends special maritime and territorial jurisdiction to
the premises of United States diplomatic, consular, military or other United States Government missions or entities in foreign States, including the buildings, parts of buildings, and land appurtenant or ancillary thereto or used for purposes of those missions or entities, irrespective of ownership....
*213 18 U.S.C. § 7(9)(A) (2006). The threshold question before us is whether this provision extends a federal court's jurisdiction to crimes committed at the Asadabad Firebase in June 2003.
When interpreting any statute, we must first and foremost strive to implement congressional intent by examining the plain language of the statute. Barnhart v. Sigmon Coal Co., 534 U.S. 438, 450, 122 S. Ct. 941, 151 L. Ed. 2d 908 (2002). "The plainness or ambiguity of statutory language is determined by reference to the language itself, the specific context in which that language is used, and the broader context of the statute as a whole." Robinson v. Shell Oil Co., 519 U.S. 337, 341, 117 S. Ct. 843, 136 L. Ed. 2d 808 (1997).
In determining the jurisdictional reach of § 7(9), the district court properly focused, at least initially, on the two critical statutory terms: "premises" and "mission." In doing so, however, the court treated "mission" as synonymous with "operation," and relied on one dictionary definition of "mission" to define the term as "a team of military specialists sent to a foreign country to assist in the training of its armed forces." See Webster's Third New International Dictionary 1445 (1993) (emphasis added). The court then found that § 7(9) provided jurisdiction over crimes committed at Asadabad in 2003 because "the United States has been conducting a military mission in Afghanistan since late 2001" and the "activities ... at the Asadabad firebase are a part of that larger mission" (emphasis added). On appeal, the Government adopts this construction, arguing that Asadabad falls within § 7(9)'s jurisdiction "[b]ecause the base was a premises for United States military operations." Brief of Appellee at 42 (emphasis added).
We believe this construction inadequate for several reasons. First, it ignores the fact that "mission" can also refer to "a permanent embassy or legation in a foreign country." Webster's, supra, at 1445 (emphasis added). This definition best harmonizes with the dictionary definition of the other critical statutory term, "premises," i.e., "a specified piece or tract of land with the structures on it." Id. at 1789. By construing § 7(9) to reach only fixed locations, rather than a mobile group of people conducting an operation, we accord importance to both "mission" and "premises," without slighting either or creating any inconsistency. See Weinberger v. Hynson, Westcott & Dunning, Inc., 412 U.S. 609, 633, 93 S. Ct. 2469, 37 L. Ed. 2d 207 (1973) (applying the "well-settled" rule of statutory construction that requires courts to give independent and internally consistent importance to each term in a statute).
Congress's placement of subsection nine within § 7 lends substantial support to this interpretation. For the other subsections in § 7 encompass only physical locations, e.g., federal property, the high seas and territorial waters of the United States, and particular types of large vehicles that themselves constitute a location. A construction of § 7(9) that adds a mobile group of people to this list of locations would seem very odd indeed.
Further, that Congress included "military... missions" in a particular list of missions i.e., "diplomatic, consular, military or other United States Government missions" indicates that Congress intended "premises of a military mission" to denote the same sort of permanent location. See United States v. Williams, ___ U.S. ___, 128 S. Ct. 1830, 1839, 170 L. Ed. 2d 650 (2008) ("[A] word is given more precise content by the neighboring words with which it is associated."). Every other *214 "mission" encompassed within this statutory list uses the word as a foreign relations term of art signifying a particular type of fixed place. See A.T. Massey Coal Co. v. Holland, 472 F.3d 148, 160 (4th Cir.2006) (discussing that context can indicate that Congress has used a term of art); see also United States v. County of Arlington, Va., 702 F.2d 485, 487 (4th Cir.1983) (noting that the Vienna Convention on Diplomatic Relations, April 18, 1961, 23 U.S.T. 3227, defines "premises of the mission" as "the buildings and lands `used for the purposes of the mission'").
Finally, although certainly not dispositive of the issue before us, § 7(9)'s legislative history also supports this construction. See Babbitt v. Sweet Home Chapter of Cmtys. for a Great Or., 515 U.S. 687, 704-08, 115 S. Ct. 2407, 132 L. Ed. 2d 597 (1995) (relying on legislative history to reinforce the Court's statutory construction). Congress enacted § 7(9) at least in part to "explicitly extend" § 7's extraterritorial reach to "U.S. diplomatic and consular premises" overseas that is, to fixed locations and not to mobile groups of people. See Administration's Draft Anti-Terrorism Act of 2001: Hearing Before the H. Comm. On the Judiciary, 107th Cong. 44 (2001). Similarly, the House Judiciary Committee report interpreted this legislation as reaching "crimes committed at United States facilities abroad." H.R.Rep. No. 107-236, at 73 (2001) (emphasis added).
For these reasons, we believe that § 7(9)'s language, context, purpose, and legislative history all indicate that "the premises of ... military ... missions" refers to fixed physical locations, i.e., land and buildings, on which the United States has established a "military mission."
We turn then to the question of whether, by June 2003, the "premises" of Asadabad constituted a United States "military mission" so as to render it within the criminal jurisdiction of a federal district court.
Clearly, long-established and permanent U.S. military bases abroad, e.g., Naval Air Facility Atsugi in Japan and Ramstein Air Base in Germany, constitute "premises" of a "military mission" under § 7(9). These bases are the straightforward "military" analogue to embassies, the "diplomatic" and "consular" missions plainly within § 7(9)'s scope. On the other hand, we doubt that § 7(9) reaches so broadly as to encompass any area that U.S. soldiers occupy, no matter how temporary or mobile their presence. For example, we agree with Passaro that § 7(9) would not reach any piece of Afghan soil on which a soldier "pitches his pup tent." Brief of Appellant at 29.
In cases that fall between these two extremes, courts must consider a number of common-sense, objective factors to determine whether a particular location qualifies as the "premises" of a United States "military mission" for purposes of § 7(9). Relevant factors include the size of a given military mission's premises, the length of United States control over those premises, the substantiality of its improvements, actual use of the premises, the occupation of the premises by a significant number of United States personnel, and the host nation's consent (whether formal or informal) to the presence of the United States. This list surely does not exhaust every factor relevant to determining § 7(9)'s reach; nor is any factor a prerequisite for jurisdiction. But these factors do bring to bear relevant, objective considerations in resolving this question.
Applying these factors to Asadabad leads us to conclude that it possesses all the qualities of a permanent U.S. military *215 base abroad, albeit on a smaller scale, and thus falls squarely within the ambit of § 7(9). First, with respect to size, Asadabad is a substantial facility, in both the area it covers and in the fortifications it includes. Although trial witnesses differed in estimates of its size, it appears that Asadabad covered approximately 350 yards on each side, or over 25 acres. In addition, a ten-foot-high and two-foot-thick heavy mud wall encircled the entire facility, making Asadabad just as discrete and set off from its surroundings as an embassy compound.
Second, as to the duration of United States control, although the United States initially took Asadabad by force, by the time Passaro arrived in May 2003, the United States had controlled Asadabad for approximately 18 months. This clearly provides objective evidence that the United States had taken permanent control of the base as of that date.
Third, by May 2003 the United States had substantially improved and fortified Asadabad. When the United States initially seized the compound, it consisted of little more than a few internal walls and a couple of rooms within the outer wall of the compound. By May 2003, however, coalition forces had improved the facility such that it contained approximately a dozen buildings, providing offices, living quarters, guard towers, and detention facilities. The Army had also installed electric generators and above-ground plumbing. Finally, building on Asadabad's already-formidable outer wall, the military had constructed extensive, multi-layered security fortifications. Such improvements indicate the permanence and substantiality of the U.S. presence at Asadabad.
Fourth, with respect to use, the United States, albeit on a smaller scale, has used Asadabad in much the same way as it has used larger military bases in Afghanistan (e.g., Bagram Air Base) and around the world. In all of these places, as in Asadabad, the United States has established significant facilities to conduct military operations, gather intelligence, and train local forces. Moreover, in Asadabad, as in these other outposts, the United States has stationed a significant military force (over 200 military and paramilitary personnel) on a long-term basis. With such forces come all of the necessary logistical support food, water, weaponry and ammunition, and sleeping facilities. Thus, the similarity between Asadabad and these larger facilities provides further objective evidence that Asadabad had become a permanent United States installation by June 2003.
Finally, we note that by June 2003, the United States and Afghanistan had entered into significant bilateral agreements. See, e.g., Agreement For the Establishment and Operation of United States Radio Transmitting Facilities in Afghanistan, U.S. AFG., Oct. 3, 2002, 2002 U.S.T. LEXIS 65. One agreement outlined the rights and obligations of United States forces in Afghanistan, effectively granting the Afghan government's imprimatur to the American presence in Afghanistan. See Agreement Regarding the Status of United States Military and Civilian Personnel of the U.S. Department of Defense Present in Afghanistan, U.S. Afg., May 28, 2003, 2002 U.S.T. LEXIS 100 [hereinafter Personnel Agreement].[2]
*216 Although we do not view such bilateral agreements as required in every case, we do view a host nation's consent to United States presence as relevant. Of course, such consent can be implicit or informal. But, in all events, bilateral agreements constitute a significant objective indication of consent. Given the fluid and chaotic context of the war in Afghanistan in 2003, these agreements provide relevant objective evidence of both the permanence of the United States presence at Asadabad and United States control of the premises.
For all of these reasons, we believe that § 7(9) extends federal criminal jurisdiction over assaults committed by U.S. nationals at Asadabad in June 2003. By that time, the United States had retained control for nearly a year and a half over this significant, discrete tract of land, maintaining a meaningful permanent presence to conduct significant military operations, which the Afghan government sanctioned. Accordingly, § 7(9) extended federal criminal jurisdiction to these premises.[3]
Having found that, by June 2003, § 7(9) provided a basis for federal jurisdiction over crimes committed at the Asadabad Firebase, we next address Passaro's contention that we should nonetheless refuse to permit § 7(9) to form the jurisdictional basis for the specific federal assault crimes charged against him. Passaro offers two theories in support of this contention; we address each in turn.
Passaro argues that the district court unconstitutionally overstepped its bounds by adjudicating the charges brought against him. He claims that allowing Article III federal courts to try his case intrudes on the Executive Branch's virtually boundless discretion in foreign affairs and war powers. This argument ignores a critical fact: the Executive itself elected to bring this prosecution.
A court does not intrude on any Executive Branch prerogative by holding that Congress empowered the Executive to prosecute criminal activity and provided a forum for adjudication of that prosecution. Passaro can cite no case holding, or even suggesting, that the exercise of prosecutorial discretion by the Executive somehow intrudes on the functions of the Executive. Instead, the cases on which Passaro relies involve private citizens suing the government in tort or in some other civil action. See, e.g., Tiffany v. United States, 931 *217 F.2d 271 (4th Cir.1991). Every case even marginally on point expressly rejects arguments like Passaro's. See, e.g., United States v. Poindexter, 725 F. Supp. 13, 23 (D.D.C.1989) ("Nothing in this case would require the Court to adjudicate foreign policy issues.... [A]ll that is involved is the question whether one particular individual... has violated the [law].").
In fact, the Supreme Court itself has sanctioned a far greater intrusion on Executive prerogatives during a criminal prosecution. In United States v. Nixon, 418 U.S. 683, 692-93, 94 S. Ct. 3090, 41 L. Ed. 2d 1039 (1974), the President contended that an "intra-branch" dispute rendered nonjusticiable his refusal to release assertedly privileged documents. The Supreme Court soundly rejected this argument, holding the controversy within Article III power because "the matter is one arising in the regular course of a federal criminal prosecution," a kind of controversy that "courts traditionally resolve." Id. at 696-97, 94 S. Ct. 3090.
If the Executive action in Nixon did not offend separation-of-powers principles, the Executive action here the Government's prosecution of Passaro surely does not. On the contrary, refusing to hear this case would significantly intrude on both the Executive and Legislative Branches by rejecting Congress's definition of criminal activity and the Executive's lawful decision to prosecute. See United States v. Funmaker, 10 F.3d 1327, 1333 (7th Cir.1993) (holding that once a court decides that a congressional enactment applies, "the judiciary must entertain all prosecutions by the executive branch undertaken pursuant to that law") (emphasis added). Just as important, accepting Passaro's argument would undermine the basis for any war crimes prosecution in an Article III court and would similarly prevent the judiciary from reviewing military prosecutions on a writ of habeas corpus. See, e.g., Burns v. Wilson, 346 U.S. 137, 142, 73 S. Ct. 1045, 97 L. Ed. 1508 (1953). Accordingly, we reject Passaro's separation-of-powers argument.[4]
Alternatively, Passaro contends that § 113 is void for vagueness as applied to him. The Supreme Court has clearly articulated the rigorous standard that criminal defendants must meet to establish such a claim. "A conviction fails to comport with due process" and so is void for vagueness only "if the statute under which it is obtained [1] fails to provide a person of ordinary intelligence fair notice of what is prohibited, or [2] is so standardless that it authorizes or encourages seriously discriminatory enforcement." United States v. Williams, ___ U.S. ___, 128 S. Ct. 1830, 1845, 170 L. Ed. 2d 650 (2008). Try as he might, Passaro has utterly failed to demonstrate that § 113 is vague in either respect.
With respect to notice, Passaro repeatedly insists that § 113, which criminalizes assault, provided him no notice that his conduct constituted assault and so violated the law. Passaro does not contend that the term "assault" in and of itself is vague. Such an argument would be doomed because courts have uniformly recognized that various federal statutes criminalizing "assault" incorporate the *218 long-established common law definition of that term. See, e.g., United States v. Chestaro, 197 F.3d 600, 604-05 (2d Cir. 1999) (rejecting contention that federal assault statute is void for vagueness); United States v. Dupree, 544 F.2d 1050, 1051-52 (9th Cir.1976) (same).
What Passaro instead contends is that § 113 provided him no notice that a "battlefield interrogation" of a terrorist suspect conducted by the CIA constituted assault. That argument fails for several reasons. First, it rests on Passaro's mischaracterization of the facts underlying the charges brought against him. No true "battlefield interrogation" took place here; rather, Passaro administered a beating in a detention cell. Nor was this brutal assault "conducted by the CIA" rather, Passaro was a civilian contractor with instructions to interrogate, not to beat. In any event, § 113 contains no exceptions for a victim's status as an alleged terrorist or a perpetrator's status as a federal agent.
Thus, although there may be cases in which § 113's application is uncertain, the facts of Passaro's case clearly fall within the scope of that statute. We therefore have no hesitation in concluding that § 113's prohibition of assault placed Passaro on notice that his repeated kicking and striking of Wali constituted criminal acts. See Parker v. Levy, 417 U.S. 733, 756, 94 S. Ct. 2547, 41 L. Ed. 2d 439 (1974) (noting that in challenges, like this one, that lie outside of the First Amendment context, a defendant "to whose conduct a statute clearly applies may not successfully challenge it for vagueness").
We further reject Passaro's contention that § 113 as applied to this "interrogation" must be void for vagueness because it would criminalize acts otherwise permissible in the heat of battle.[5] To accept this argument would equate a violent and unauthorized "interrogation" of a bound and guarded man with permissible battlefield conduct. To do so would ignore the high standards to which this country holds its military personnel.
Passaro's notice argument, rather than establishing vagueness, merely constitutes an attempt to justify his crimes. Passaro raised this and similar justification defenses before the jury. The jury rejected them, and Passaro does not contend that the jury's verdict fails for lack of evidence. We cannot and will not permit him to transform a justification defense into a constitutional vagueness challenge.
Passaro's second vagueness argument that the Government's assertedly selective application of § 113 to him renders the statute unconstitutionally void for vagueness also fails. Stripped of its constitutional garb, this argument merely constitutes Passaro's objection to the Government's decision to prosecute him for violations of § 113.
*219 This objection is predictable. Indeed, criminal defendants commonly complain that other persons are more worthy of criminal prosecution. But our constitutional system leaves it to the discretion of the Executive Branch to decide who will face prosecution. See United States v. Armstrong, 517 U.S. 456, 464, 116 S. Ct. 1480, 134 L. Ed. 2d 687 (1996). Unless a defendant provides "clear evidence" to overcome the presumption that a government prosecutor has acted lawfully and without discrimination a "particularly demanding" standard he cannot demonstrate a constitutional violation for selective prosecution. Reno v. Am.-Arab Anti-Discrimination Comm., 525 U.S. 471, 489, 119 S. Ct. 936, 142 L. Ed. 2d 940 (1999).
Passaro fails to cite any evidence that in prosecuting him for violations of § 113, the Government unlawfully or discriminatorily exercised its prosecutorial discretion. Cf. Yick Wo v. Hopkins, 118 U.S. 356, 373-74, 6 S. Ct. 1064, 30 L. Ed. 220 (1886). Accordingly, he provides no basis for judicial interference with that discretion, especially in light of the difficulty of such decisions in this sensitive diplomatic and military context.[6]
Having assured ourselves that the district court properly exercised jurisdiction over the Government's prosecution of Passaro, we turn to Passaro's remaining challenges to his conviction.
Passaro argues that the Government and the district court used the Classified Information Procedures Act (CIPA), Pub.L. No. 96-456, 94 Stat. 2025 (1980) (codified at 18 U.S.C.App. 3 §§ 1-16 (2006)), to deny him a fair trial. Specifically, Passaro argues that the district court permitted Government "misuse" of CIPA that (1) prevented him from "obtaining or presenting" evidence necessary to his defense and (2) allowed the Government to present "distorted" information regarding the circumstances surrounding his assault of Wali that "prevent[ed] proper jury instructions." Brief of Appellant at 61, 63. We review a court's CIPA rulings for an abuse of discretion, granting the court wide discretion in handling classified evidence. United States v. Abu Ali, 528 F.3d 210, 247, 253 (4th Cir.2008).
We begin with Passaro's objections to the limits the district court imposed on disclosure of classified information. CIPA provides a procedural framework by which a court balances the defendant's interest in a fair trial and the Government's interest in protecting national security information. United States v. Dumeisi, 424 F.3d 566, 578 (7th Cir.2005). CIPA permits a court to review classified evidence in camera to determine its relevance and admissibility, and then order disclosure (with or without redaction) of the evidence or issue protective orders against disclosure. 18 U.S.C.App. 3 §§ 3, 5-6 (2006); United *220 States v. Smith, 780 F.2d 1102, 1105-06 (4th Cir.1985) (en banc). CIPA does not, however, alter the substantive rules of evidence, including the test for relevance: thus, it also permits the district court to exclude irrelevant, cumulative, or corroborative classified evidence. Id. at 1106, 1110; see also Fed.R.Evid. 403.
Here the district court conducted the required CIPA hearing as to each of Passaro's requests for classified information disclosure and issued a detailed order on the admissibility of each. See 18 U.S.C.App. 3 § 6. The court admitted some of the evidence in full, admitted some in redacted form, and excluded some as irrelevant, cumulative, or corroborative.
With one exception, Passaro does not object to the redaction or exclusion of any specific piece of evidence.[7] Rather, he makes unsupported assertions, which amount to little more than insinuation of error. Having reviewed the record ourselves, we agree with the Government that the district court's rulings allowed Passaro to present a full defense to the jury regarding his conduct and the circumstances surrounding that conduct. Thus, we find no abuse of discretion in the district court's limitations on the introduction of classified evidence.
Nor can we find fault with the district court's orders withholding from Passaro discovery of certain classified information.
Passaro argues at length that the district court permitted the Government to use CIPA as a "sword" to prevent him from discovering the authorization necessary to his public authority defense. But to establish such an affirmative defense, Passaro must prove that someone with actual authority sanctioned an otherwise unlawful act. United States v. Fulcher, 250 F.3d 244, 254 (4th Cir.2001). Passaro offered no showing that such evidence exists. We recognize that such a showing may well be difficult given national security concerns, but at the very least Passaro could have proffered a specific conversation that he had with a superior, or a particular document on which he relied, that purported to authorize his brutal "interrogation" of Wali. Passaro made no such proffer.
On the other hand, the Government presented substantial evidence that Passaro never received authorization for his interrogation methods. Passaro's two direct superiors at Asadabad unequivocally testified at trial (and thus were subject to cross-examination by Passaro's counsel) that they never authorized Passaro's actions. Another civilian contractor, who had undergone training with Passaro, similarly testified that the CIA did not permit interrogators to strike detainees. Finally, the Government offered a summary of CIA interrogation policy. The district court, after examining ex parte the full top-secret policy, admitted a redacted summary into evidence pursuant to 18 U.S.C.App. 3 § 4. This document supports the testimony of the Government witnesses and offers no support to an affirmative public authority defense for Passaro's acts.
*221 Nevertheless, Passaro contends that the district court abused its discretion when it quashed his subpoenas to CIA officials who he asserts could have provided support for his public authority defense. To obtain such compulsory process of a witness, the Sixth Amendment requires a defendant to demonstrate that the witness will testify "`in his favor.'" United States v. Moussaoui, 382 F.3d 453, 471 (4th Cir.2004) (quoting U.S. Const. amend. VI). Passaro has failed to make any showing that the testimony of additional CIA officials would have aided his defense and, in fact, the testimony of his two superiors and the written CIA policy indicate that any additional subpoenaed witness would not have testified "in his favor." Moreover, Passaro admits that he did not comply with the regulatory procedures governing requests for information from a CIA employee. 32 C.F.R. §§ 1905.3-.4 (2008); see also United States ex rel. Touhy v. Ragen, 340 U.S. 462, 71 S. Ct. 416, 95 L. Ed. 417 (1951). We see no reason to permit Passaro to depart from these procedures.
For all these reasons, we reject Passaro's CIPA arguments.
In addition to his CIPA arguments, Passaro also maintains that the district court committed reversible error in rejecting his proposed jury instruction regarding justifiable use of reasonable force. A district court commits reversible error in refusing to provide a proffered jury instruction only when the instruction "(1) was correct; (2) was not substantially covered by the court's charge to the jury; and (3) dealt with some point in the trial so important, that failure to give the requested instruction seriously impaired the defendant's ability to conduct his defense." United States v. Lewis, 53 F.3d 29, 32 (4th Cir.1995) (internal quotation marks omitted). We review the district court's decision to give or refuse to give a jury instruction for abuse of discretion. United States v. Moye, 454 F.3d 390, 397-98 (4th Cir.2006). Moreover, "we do not view a single instruction in isolation; rather we consider whether taken as a whole and in the context of the entire charge, the instructions accurately and fairly state the controlling law." United States v. Rahman, 83 F.3d 89, 92 (4th Cir.1996).
Analogizing to a police arrest of a dangerous person, Passaro contends that the court should have instructed the jurors that they could excuse his conduct if they found that he, a "federal officer," had used a level of force reasonably necessary to effectuate and maintain control of a criminal. Specifically, Passaro's proposed instruction required the jury to consider, inter alia:
[T]he circumstances which led the United States forces to take [Wali] into custody [and] the information Mr. Passaro possessed about [Wali], to include information linking [Wali] to terrorist groups and to recent and frequent ambushes....
We reject the notion that Passaro, a CIA contractor charged with questioning a man who was already detained and shackled, could avail himself of an excessive force defense reserved for law enforcement officers attempting to subdue or restrain a threatening person. See, e.g., Waterman v. Batton, 393 F.3d 471, 476-77 (4th Cir.2005). Moreover, Passaro's proffered instruction is not legally correct because it improperly directs the jury to consider irrelevant factors demonstrating Wali's alleged past history of general dangerousness. This is not the standard. Rather, a fact finder assesses the reasonableness of an officer's use of force to effectuate or maintain control of a detainee on the basis of the detainee's conduct in *222 the moments surrounding the officer's use of force. Cf. Graham v. Connor, 490 U.S. 386, 396, 109 S. Ct. 1865, 104 L. Ed. 2d 443 (1989).
To whatever extent Passaro's proposed instruction is relevant, the district court's charge to the jury taken as a whole sufficiently accounted for it. For instance, the court instructed the jury that Passaro lacked the requisite criminal intent if he "struck Abdul Wali in order to achieve another objective and not with the express intent to cause bodily harm."
Thus, the district court did not abuse its discretion in refusing to give Passaro's proposed jury instruction.
Finally, we turn to the sentencing issues. The district court calculated Passaro's offense level as 23, after finding a base offense level of 14, see U.S. Sentencing Guidelines Manual § 2A2.2(a) (2008), and applying enhancements for threatened use of a dangerous weapon (three levels), id. § 2A2.2(b)(2)(C), infliction of serious bodily injury (four levels), id. § 2A2.2(b)(3)(D), and involvement of a vulnerable victim (two levels), id. § 3A1.1(b)(1). With Passaro's criminal history category of I, this established an advisory guidelines range of 46-57 months. From this range, the district court departed upward six levels for "extreme conduct," see id. § 5K2.8, resulting in an offense level of 29 and an adjusted advisory Guidelines range of 87-108 months. The district court then sentenced Passaro to a 100-month term of imprisonment.
Passaro and the Government, albeit for different reasons, agree that the district court erred in applying the three-level enhancement for threatened use of a "dangerous weapon." Id. § 2A2.2(b)(2)(C). Although not clear from the record, it appears that the district court based this enhancement on Passaro's use of a heavy flashlight or his shod foot to beat Wali.
Passaro argues that the district court erred because the jury found that Passaro had actually kicked Wali with his foot, rather than threatening to kick him. Thus, Passaro contends, the evidence in no event supports a finding that Passaro merely threatened to kick Wali. According to Passaro, the district court therefore should have either applied the four-level enhancement for actual use of a dangerous weapon, id. § 2A2.2(b)(2)(B), or no enhancement at all. The Government agrees that the district court erred, but only because the court should have applied the four-level enhancement for actual use of a dangerous weapon.
The Guideline-sanctioned definition of dangerous weapon encompasses an extremely broad range of instrumentalities. Id. §§ 2A2.2 cmt. n. 1, 1B1.1 cmt. n. 1(D). This definition allows a trier of fact to consider as a dangerous weapon a knife, gun, shoe, dog, rake, or any other item adapted to causing death or serious bodily injury. See United States v. Dayea, 32 F.3d 1377, 1379 & n. 2 (9th Cir.1994) (collecting cases). Although at sentencing the parties and court engaged in much discussion of the boot and the flashlight, the court ultimately applied the dangerous weapon enhancement with no explanation. We agree with Passaro that the district court must explicitly find by a preponderance of the evidence what, if any, instrumentality constituted the basis for the dangerous weapon enhancement. If the district court finds that a specific instrumentality constituted a dangerous weapon and that Passaro actually used that weapon, then the Guidelines direct that the court apply the four-level enhancement for actual use of a dangerous weapon. Thus, *223 we must remand for further findings on this issue.
In addition, both parties agree that the district court erred in imposing a six-level upward departure. In imposing the departure, the court cited § 5K2.8 of the Guidelines, which provides for an upward departure when the defendant's conduct "was unusually heinous, cruel, brutal, or degrading." U.S. Sentencing Guidelines Manual § 5K2.8 (2008). But, as the Government acknowledges, the district court offered no rationale for concluding that § 5K2.8 justified the chosen upward departure.
It is axiomatic that a district court commits reversible procedural error when it fails to explain a departure or variance, whether upward or downward. United States v. Carter, 564 F.3d 325, 328 (4th Cir.2009). Moreover, without such an explanation we cannot determine whether a departure was substantively reasonable. Id. Thus, we must vacate the sentence and remand for resentencing, at which the district court can explain fully the departure, if any, that it chooses to impose.
Passaro asserts additional sentencing errors, alleging that the district court erred in imposing a vulnerable victim enhancement, U.S. Sentencing Guidelines Manual § 3A1.1(b)(1) (2008), and a four-level increase for serious bodily injury, id. § 2A2.2(b)(3)(D). The two errors discussed above, however, already require a remand for resentencing, so we need not address these additional sentencing issues. In all events, nothing herein should be read to preclude the district court from reconsidering its earlier findings.
This case has required us to determine whether a federal court has jurisdiction over the trial of an American citizen for committing brutal assaults abroad. Congress has determined that individuals committing such crimes on the premises of United States military missions abroad are subject to prosecution in United States federal courts. The Executive has determined to bring the first such case against David Passaro. We are satisfied that Passaro received a fair trial from a conscientious jury, in a court that had jurisdiction to try him. Thus, after careful consideration of the challenges Passaro presents, we affirm Passaro's conviction in all respects. In light of the conceded sentencing errors, however, we vacate Passaro's sentence and remand for resentencing.
AFFIRMED IN PART AND VACATED AND REMANDED IN PART.
[1] Because Wali's family removed his body from Asadabad after his death and refused to allow an autopsy, the Government explains that it lacked evidence as to the cause of Wali's death or a basis for any other criminal charges.
[2] Relying on the treaty exclusion in § 7(9) ("Nothing in this paragraph shall be deemed to supersede any treaty or international agreement with which this paragraph conflicts."), Passaro contends that the absence of a specific jurisdictional grant in this Personnel Agreement precludes American jurisdiction over American civilians in Afghanistan. But the Agreement, which covers both military and civilian personnel, specifically does grant the United States "criminal jurisdiction over United States personnel" in Afghanistan. See Personnel Agreement, 2002 U.S.T. LEXIS 100, at *4. Similarly, contrary to Passaro's suggestion, the Vienna Convention does not preclude the application of § 7(9) to Passaro. The Vienna Convention regulates and extends certain privileges to the premises of diplomatic missions; it simply does not address the kind of military mission at issue in this case.
[3] For these same reasons, we cannot conclude, as Passaro contends, that § 7(9) contains such "grievous ambiguity or uncertainty" as to invoke the rule of lenity. Muscarello v. United States, 524 U.S. 125, 138-39, 118 S. Ct. 1911, 141 L. Ed. 2d 111 (1998) (internal quotation marks omitted). We also reject Passaro's argument that § 7(9) does not apply to him because of the canon against extraterritoriality. See EEOC v. Arabian Am. Oil Co., 499 U.S. 244, 248-59, 111 S. Ct. 1227, 113 L. Ed. 2d 274 (1991), superseded by statute, Civil Rights Act of 1991, Pub.L. No. 102-166, § 109, 105 Stat. 1071, 1077. The statute expressly applies to "premises ... in foreign States," 18 U.S.C. § 7(9) (emphasis added); thus, Congress intended § 7(9) to reach extraterritorially. See Arabian Am. Oil Co., 499 U.S. at 248, 111 S. Ct. 1227 (noting that the canon against extraterritoriality applies only when congressional intent is "unexpressed").
[4] Passaro also contends that the constitutional avoidance canon prevents the Government from prosecuting him, asserting that we should construe § 7(9) to avoid separation-of-powers concerns. See INS v. St. Cyr, 533 U.S. 289, 299-300, 121 S. Ct. 2271, 150 L. Ed. 2d 347 (2001). We reject this contention for the same reasons that we reject Passaro's separation-of-powers argument.
[5] Passaro attempts to bolster this argument with a 2003 Office of Legal Counsel (OLC) memorandum, since retracted, which he maintains proves that the Department of Justice believed § 113 did not apply to battlefield interrogations. See Memorandum for William J. Haynes II, Gen. Counsel of the Dep't of Def. (March 14, 2003). This memorandum recognized, as we do, that some foreign military locations would not fall within the "special maritime and territorial jurisdiction of the United States." Id. at 21. But because the memorandum clearly states that § 7 reaches "U.S. military bases in foreign states," id. at 19-20, and because Asadabad is such a military base, the memorandum undermines, rather than supports, Passaro's position. Indeed, the memorandum addresses at length limitations that § 113 places on interrogations within the special maritime and territorial jurisdiction of the United States. Id. at 24-30.
[6] Passaro's contention that § 7(9), either alone or taken together with § 113, renders his prosecution void for vagueness also fails. Simply because this case constitutes the first prosecution under § 7(9) of the acts outlawed by § 113 does not render § 7(9) unconstitutionally vague. See United States v. Buculei, 262 F.3d 322, 333 (4th Cir.2001) ("[I]t is irrelevant that [the defendant's] prosecution under this statute is `a novel construction,' or that it is the first time the Government has proceeded under this theory."). As we have explained at length, § 7(9)'s "premises of ... military ... missions" unambiguously encompasses permanent U.S. military outposts abroad, such as Asadabad.
[7] The lone exception is the 2003 OLC Memorandum on which Passaro relies throughout. See supra note 5. But the record lacks any evidence that Passaro read or knew of, let alone relied on, this memorandum prior to his assault on Wali. Even if he had, such reliance would simply amount to a mistake of law, which provides no defense to the assault charges. See Cheek v. United States, 498 U.S. 192, 199, 111 S. Ct. 604, 112 L. Ed. 2d 617 (1991). For similar reasons, CIA memoranda that have recently come to light do not aid Passaro's defense. See, e.g., Mark Mazzetti & Scott Shane, Debate Over Interrogation Methods Sharply Divided the Bush White House, N.Y. Times, May 4, 2009, at A13.
