Pursuant to the bilateral Extradition Treaty between the United States and Peru, see Extradition Treaty, U.S.-Peru, July 26, 2001, S. Treaty Doc. No 107-6 ("Extradition Treaty"), the United States Government seeks to extradite Wilmer Yarleque Ordinola, a Peruvian national, to Peru so that he can stand trial for the alleged crimes of aggravated homicide, aggravated kidnapping, forced disappearance of persons, inflicting major intentional injuries, and delinquent association. Before a magistrate judge, Ordinola claimed, inter alia, that he was not extraditable because his alleged crimes fit within the Treaty's exception for political offenses. The magistrate judge, acting as the finder of fact, disagreed, finding that Ordinola's offenses could not be defined as political for purposes of the Treaty. Ordinola filed a petition for a writ of habeas corpus, and the district court granted the writ, determining that the political offense exception barred extradition in this case. The Government appealed. For the reasons identified below, we conclude that the magistrate judge did not err in finding Ordinola extraditable. We therefore vacate the district court's grant of the writ and remand for reentry of a Certification of Extraditability.

In the late 1960s, Abimael Guzman, a Peruvian philosophy professor and leader of a faction of Peru's Communist Party, founded the radical left-wing guerrilla movement known as Sendero Luminoso, or "the Shining Path." The Shining Path has since been described as a "highly organized guerrilla organization with a Maoist communist ideology dedicated to the violent overthrow of Peru's democratic government and social structure." Sotelo-Aquije v. Slattery, 17 F.3d 33, 35 (2d Cir. 1994). By the mid-1980s, the Shining Path had grown from a relatively small regional movement into a much larger and more dispersed one that engaged in numerous acts of violence against Peruvian government officials, peasants, journalists, professors, teachers, and others. At its height, the Shining Path was characterized as "[s]ophisticated, well-organized," and "probably the most brutal, vindictive, and elusive terrorist insurgency in the Western Hemisphere." (J.A. at 138 (The Threat of the Shining Path to Democracy in Peru: Hearing Before the Subcomm. on Western Hemisphere Affairs of the H. Comm. on Foreign Affairs, 102d Cong. 47 (1992) (statement of Gabriela Taranzona-Sevillano, Visiting Prof. of Int'l Studies, Davidson Coll.)).)

In 1990, newly elected Peruvian president Alberto Fujimori enlisted Vladimiro Lenin Montesinos, former Advisor to the Peruvian National Intelligence Service, as a military advisor. Together, Fujimori and Montesinos provided the military with vast power and resources to fight the Shining Path,1 responding to the group's insurgency with tactics that nearly matched the Shining Path's in brutality. The result was Fujimori's creation of a military establishment that was described as "institutionally corrupt," clouded by "the worst human rights record in the hemisphere," and characterized by "a persistent record of indiscriminate action against rural populations" stemming from a "belief that popular terror, in and of itself, is a recipe for success [against the Shining Path]." (J.A. at 110, 112 (The Threat of the Shining Path to Democracy in Peru: Hearing Before the Subcomm. on Western Hemisphere Affairs of the H. Comm. on Foreign Affairs, 102d Cong. 19, 21 (1992) (statement of Gordon H. McCormick, senior social scientist, RAND Corp.)).)

During 1991, in connection with this strategy, Fujimori created a special operations paramilitary squad known as the Grupo Colina, or "Colina Group," and commissioned the group to combat the Shining Path.2 Ordinola, a veteran of the Peruvian army with experience in military intelligence, was assigned to the group and later became "group chief," meaning that he was coordinator for one of the Colina Group's three subgroups.

Peru's extradition request alleges that Ordinola, while serving in the Colina Group, kidnapped and murdered noncombatant civilians on four separate occasions in 1991 and 1992. The allegations related to these four incidents are recounted below.

1. The Barrios Altos Case.

On the evening of November 3, 1991, two trucks containing Colina Group agents, including Ordinola, traveled to the Barrios Altos neighborhood in Lima, Peru, where a pollada—a social function typically held for fund-raising purposes—was taking place in the yard of an apartment complex. Dressed in ski masks and carrying machine guns with silencers, Ordinola and around ten other agents left the trucks, entered the apartment complex yard, and began shooting at the people attending the pollada. The agents killed fifteen people, including an eight-year-old boy, and seriously injured four others.3

Julio Chiqui Aguirre, a Colina Group agent who participated in the attack, identified Ordinola as an active participant in the shootings and as the killer of the eight-year-old boy. In connection with the massacre, Peru has charged Ordinola with Aggravated Homicide, Inflicting Major Intentional Injuries, and Delinquent Association, in violation of Articles 108, 121, and 317 of the Peruvian Criminal Code.

2. La Cantuta Case.

On July 17, 1992, Ordinola and other Colina Group agents traveled to La Cantuta University. The agents, who were armed, wearing ski masks, and carrying shovels and lime,4 entered the student dormitories and forced approximately fifteen students outside. One of the agents singled out nine of the fifteen students because they allegedly had terrorist connections. The agents also forced a university professor out of his home because of an alleged link to terrorism.

Ordinola and the other Colina Group agents loaded the professor and nine students into several trucks and drove them to a location on the Ramiro Priale Highway. There, the agents made them dig trenches, forced them into the holes, shot and killed them, and buried them. One or two days later, several agents, including Ordinola, exhumed the bodies, moved them to another location, and incinerated them because they were concerned that the bodies might be discovered.5 For this incident, Peru has charged Ordinola with Aggravated Homicide, Aggravated Kidnapping, and Forced Disappearance of Persons, in violation of Articles, 108, 152, and 320 of the Peruvian Criminal Code.

3. The Barrientos Case.

On May 1, 1992, Peruvian Army General Juan Rivero Lazo summoned leaders of the Colina Group to a meeting with Jorge Fung Pineda, a private property owner who had friends in the Peruvian military. Pineda told the agents that some of his workers at his cotton mill were demanding higher wages and improved machinery. Pineda asked the agents to link the workers to terrorism and to "teach them a good lesson." (J.A. at 23.)

Later that day, Ordinola and other Colina Group agents, while dressed as civilians, armed themselves and kidnapped nine of the workers, taking one person from the road while he was riding a bicycle and the other eight from their homes. Ordinola and the other agents killed all nine people and buried them at a nearby farm. Before they left, the agents painted "written marks and phrases regarding the `Shining Path' with red paint" around the scene to make it appear that the crimes were perpetrated by the Shining Path. (J.A. at 9, 24.) For this incident, Peru has charged Ordinola with Aggravated Homicide and Aggravated Kidnapping, in violation of Articles 108 and 152 of the Peruvian Criminal Code.

4. The Bustamante Case.

In June 1992, Pedro Yauri Bustamante was living in Huacho, Peru, and working as director of the daily news program called, "Punto Final," which aired on a local radio station. On the show, Bustamante and his call-in listeners frequently criticized the Fujimori government. The Fujimori administration had classified Bustamante as a subversive activist because he had previously been investigated for terrorist activities.

On June 22, 1992, Major Santiago Martin Rivas called members of the Colina Group, including Ordinola, to a meeting and told them that the next day they would carry out an operation in Huacho. On June 23, the agents, armed and carrying shovels, lime, and ski masks, left the grounds of the former Peruvian Intelligence Service and traveled in two vehicles to an area near the beach. Major Rivas directed one group of agents to stay at the beach and dig a grave in the sand and directed the other group to kidnap Bustamante and bring him back to the beach.

Around 2:00 a.m. on June 24, 1992, the assigned group broke into Bustamante's house, tied up him and his father, confiscated several documents, and forced Bustamante into the agents' vehicle. The agents drove him to the beach, where Major Rivas interrogated Bustamante. Bustamante refused to answer Rivas's questions because he knew they were going to kill him anyway. The group forced Bustamante to dig the rest of his grave, and Agent Luis Ortiz Mantas fatally shot Bustamante in the head. The agents then buried Bustamante. Agent Aguirre identified Ordinola as a participant in Bustamante's abduction and murder. For this incident, Peru has charged Ordinola with Aggravated Homicide and Aggravated Kidnapping, in violation of Articles 108 and 152 of the Peruvian Criminal Code.

Ordinola entered the United States on February 20, 2001, and filed a petition for political asylum on June 22, 2001. On November 12, 2003, pursuant to the bilateral Extradition Treaty with the Republic of Peru, the United States Department of State received four Diplomatic Notes from the Peruvian Embassy requesting that United States authorities obtain provisional arrest warrants in connection with the extradition of Ordinola.6

On September 24, 2004, a United States magistrate judge issued a provisional arrest warrant for Ordinola pending his extradition. On June 10, 2005, the magistrate judge, having previously held an extradition hearing pursuant to 18 U.S.C.A. § 3184 (West 2000) and the terms of the Treaty, found that probable cause existed to believe that Ordinola had committed the alleged crimes. Turning to the question of whether Ordinola's crimes were extraditable under the Treaty, the magistrate judge first found that the charged crimes satisfied the dual criminality requirement.7 The magistrate judge next concluded that Ordinola's alleged crimes did not fit within the political offense exception to extradition. The magistrate judge ruled that although Ordinola's alleged crimes occurred during a severe political uprising, the crimes were not sufficiently incidental to the uprising and thus did not fall within the exception. He reasoned that any political intentions Ordinola had in committing the alleged crimes were not enough to render the acts political offenses in light of the fact that Ordinola committed the crimes against noncombatant civilians and engaged in acts that violated the laws of armed conflict and international standards of civilized conduct. Accordingly, the magistrate judge certified Ordinola's extradition to the Secretary of State.

On July 20, 2005, Ordinola filed a petition for a writ of habeas corpus in the Eastern District of Virginia, alleging that he was being held in violation of the Extradition Treaty. On November 21, 2005, the district court granted the writ, concluding that the magistrate judge had properly found probable cause to believe that Ordinola committed the charged crimes but had wrongly concluded that Ordinola's offenses did not fit within the political offense exception. The court concluded that Ordinola's crimes were sufficiently incidental to the political uprising because the Peruvian government led Ordinola to believe that the victims of Ordinola's crimes were terrorists. As such, the court reasoned that Ordinola did not knowingly kill innocent civilians.

The Government timely appealed. We have jurisdiction under 28 U.S.C.A. § 1291 (West 2006).

The political offense exception to extradition forbids countries from extraditing people who are accused of offenses that are "political" in nature. Like the vast majority of modern-day extradition treaties, the extradition Treaty between the United States and Peru provides a political offense exception:

Extradition shall not be granted if the offense for which extradition is requested constitutes a political offense.

Extradition Treaty, art. IV, sec. 2. Unfortunately, however, these treaties do not define "political offense." Accordingly, we are forced to rely on judicial constructions, history, purpose, and State Department interpretations to determine the phrase's meaning. See United States v. Al-Hamdi, 356 F.3d 564, 570 (4th Cir.2004) ("When the text [of the treaty] is ambiguous or unclear, we turn to nontextual sources for guidance."). Although the exception has been ingrained in our country's extradition treaties for well over a century, this Circuit has never expressly addressed the exception. We do so now, beginning with an explanation of the exception's history and purpose.

Extradition requests are of ancient origin.8 Extradition was the process by which states requested the surrender of "pure" political offenders, i.e., those accused of treason and contemptuous behavior toward the monarch. M. Cherif Bassiouni, International Extradition: United States Law and Practice 594 (4th ed.2002). The first known extradition treaty—representing "one of the oldest documents in diplomatic history"—was entered into by the Egyptians and Hittites circa 1280 B.C. Id. at 32. The political offense exception to extradition, however, is a far more recent development, tracing its beginnings to the Enlightenment ideals encapsulated in the French and American revolutions.

These ideals supported a belief that people possessed an inalienable right to resist and abolish tyrannical governments. See, e.g., The Declaration of Independence para. 1 (U.S.1776) ("We hold these truths to be self-evident. . . . That whenever any Form of Government becomes destructive of these ends, it is the Right of the People to alter or to abolish it . . . ."). It was with these ideals in mind that Thomas Jefferson, as Secretary of State in 1792, recommended against entering into an extradition treaty with Spain:

[M]ost Codes extend their definitions of treason to acts not really against one's country. They do not distinguish between acts against the government and acts against the Oppressions of the government. The latter are virtues: yet have furnished more victims to the Executioner than the former. Because real treasons are rare: Oppressions frequent. The unsuccessful Struggles against Tyranny have been the chief Martyrs of Treason laws in all countries. . . . [W]e should not wish then to give up to the Executioner the Patriot who fails, and flees to us.

Letter from Thomas Jefferson to William Carmichael and William Short (Apr. 24, 1792), in 6 The Works of Thomas Jefferson (Correspondence 1789-1792), at 447-48 (Paul Leicester Ford ed., 1904) (emphasis in original).

The theory underpinning the political offense exception, then, is as old as our country.9 The exception was deemed necessary to protect those people who justly fought back against their government oppressors to secure political change. In 1843, a decade after nations such as Belgium, France, and Switzerland included the political offense exception in their extradition treaties, the United States followed suit. Bassiouni, supra, at 599-600.

Traditionally, there have been two categories of political offenses: "pure" and "relative." The core "pure" political offenses are treason, sedition, and espionage. Vo v. Benov, 447 F.3d 1235, 1241 (9th Cir.2006). "Pure" political offenses do not have any of the elements of a common crime because "[s]uch laws exist solely because the very political entity, the state, has criminalized such conduct for its self-preservation." Bassiouni, supra, at 604. Such crimes are perpetrated directly against the state and do not intend to cause private injury. Most extradition treaties preclude extradition for "pure" political offenses. "Relative" political offenses, on the other hand, are common crimes that are so intertwined with a political act that the offense itself becomes a political one. Id. at 607-08. As evidenced by this discussion, while "pure" political offenses are easy to identify, determining whether a common offense is "relatively" political requires close attention to the specific facts at issue.

Most American courts addressing "relative" political offenses have developed a two-prong test to determine whether an offense is sufficiently political to fall within the exception. Known as "the incidence test," it asks whether (1) there was a violent political disturbance or uprising in the requesting country at the time of the alleged offense, and if so, (2) whether the alleged offense was incidental to or in the furtherance of the uprising. See, e.g., Vo, 447 F.3d at 1241. We, too, adopt the incidence test as our lodestar.

Pursuant to 18 U.S.C.A. § 3184, a magistrate judge has jurisdiction to review the evidence to determine whether an extradition request can be sustained under a treaty.10 If the magistrate determines that the suspect is extraditable, it must then certify the question to the Secretary of State, who is entrusted with the ultimate executive power of determining extradition. 18 U.S.C.A. § 3186 (West 2000). Thus, before extradition can take place, an applicable treaty must exist, and the suspect must be able to challenge application of the treaty by both judicial and executive review.

In Ornelas v. Ruiz, 161 U.S. 502, 16 S. Ct. 689, 40 L. Ed. 787 (1896), the Supreme Court explained that the political offense extradition question presents "a question of mixed law and fact, but chiefly of fact." Id. at 509. The Court, however, afforded broad deference to the magistrate's finding of extraditability, explaining that it was reviewing to determine only whether the magistrate "had no choice, on the evidence, but to hold . . . that this was a movement in aid of a political revolt . . . and that acts which contained all the characteristics of crimes under the ordinary law were exempt from extradition because of the political intention of those who committed them." Id. at 511, 16 S. Ct. 689 (emphasis added).

This "no choice" language brings to mind the substantial evidence or clear error standard employed by courts today. For example, in deciding under the substantial evidence standard whether to reverse the immigration agency's decision to deny asylum,11 we will uphold the decision unless the petitioner can show that the evidence he presented would have "compelled" a reasonable factfinder to conclude otherwise. See, e.g., I.N.S. v. Elias-Zacarias, 502 U.S. 478, 483-84, 112 S. Ct. 812, 117 L. Ed. 2d 38 (1992). Ornelas's standard of review also meshes nicely with the language of § 3184, which instructs the factfinder to sustain the charge under the provisions of the treaty if "he deems the evidence sufficient." 18 U.S.C.A. § 3184.

Thus, we are left with a standard of review that recognizes the political offense question as a mixed question of law and fact, but mostly of fact. Ornelas, 161 U.S. at 509, 16 S. Ct. 689. On habeas review, we must grant a magistrate's factual findings great deference and affirm the decision unless it is "palpably erroneous in law" and a reasonable factfinder would have had "no choice" but to conclude that the offender was acting in furtherance of a political uprising. Id. at 509, 511, 16 S. Ct. 689.

The Government argues that the district court failed to afford the magistrate judge's opinion proper deference and that it misapplied the political offense exception because Ordinola's offenses were not incident to the political uprising in Peru.12 It contends that if allowed to stand, the opinion would extend the narrow doctrine "to virtually every misdeed, no matter how heinous, committed by a person purporting to act under a political cloak." (Appellant's Br. at 29.) Ordinola, on the other hand, argues that the district court did not err and that there exists a close nexus between Ordinola's alleged actions and his political objective. We agree in principle with the Government.

There is no direct appeal for an individual found to be extraditable by a magistrate. See Collins v. Miller, 252 U.S. 364, 369, 40 S. Ct. 347, 64 L. Ed. 616 (1920). Rather, a writ of habeas corpus is the only available means to challenge the magistrate's finding. A judge conducting habeas review of an extradition case is subject to substantial limitations and is not free "to rehear what the magistrate has already decided." Fernandez v. Phillips, 268 U.S. 311, 312, 45 S. Ct. 541, 69 L. Ed. 970 (1925). The magistrate's decision therefore "cannot be reviewed on the weight of the evidence." Ornelas, 161 U.S. at 509, 16 S. Ct. 689. "[H]abeas corpus is available only to inquire whether the magistrate had jurisdiction, whether the offense charged is within the treaty and, by a somewhat liberal extension, whether there was any evidence warranting the finding that there was reasonable ground to believe the accused guilty" of the asserted crimes. Fernandez, 268 U.S. at 312, 45 S. Ct. 541. Under habeas review, "the political offense question is reviewable . . . as part of the question of whether the offense charged is within the treaty." Quinn v. Robinson, 783 F.2d 776, 791 (9th Cir.1986).

As an initial matter, then, the district court plainly erred in reviewing de novo the entirety of the magistrate judge's decision and affording his factual findings no deference. For one of many examples, the district court's finding that "Ordinola did not knowingly murder innocent civilians," (J.A. at 528), is contradicted by the magistrate judge's opposite finding that Ordinola's victims "were clearly civilians." (J.A. at 31.) As another example, the magistrate judge found that the young boy allegedly murdered by Ordinola "as he was running to help his dying father was almost certainly not a terrorist." (J.A. at 31.) The district court, however, rejected this finding as "irrelevant" because of its independent factual finding that "several of the Shining Path members themselves barely reached adolescence before participating in the insurgency's violent and brutal tactics." (J.A. at 527.)

Such rejection of the magistrate judge's factfinding was improper because the district court was not free to "to rehear what the magistrate has already decided." Fernandez, 268 U.S. at 312, 45 S. Ct. 541. Rather, as a habeas court, the district court was constrained by the same standard of review that is articulated in the discussion above for this Court. Because the magistrate judge was the factfinder and its decision was not subject to direct appeal, the district court erred by not paying substantial deference to the magistrate judge's factual findings. See, e.g., Quinn, 783 F.2d at 792 (explaining that "[i]t would make little sense for us to ignore the factual findings of the judicial tribunal that made the initial factual determinations and defer, instead, to the differing factual findings made by a [later] tribunal that merely reviewed the record of the earlier proceedings"); Eain v. Wilkes, 641 F.2d 504, 508 (7th Cir.1981) ("The scope of habeas corpus review in extradition cases is a limited one, according due deference to the magistrate's initial determination."). Thus, although the district court was free to make its own legal conclusions so long as they were supported by the magistrate judge's factual findings, it was not free to ignore or misinterpret those findings in an effort to reach a desired legal conclusion.

To fall within the political offense exception, Ordinola's alleged actions must have been incidental to or in furtherance of a violent political uprising in Peru. Although Ordinola's actions occurred in the course of a violent political uprising, he cannot show that the magistrate judge erred in finding that those actions were not in furtherance of quelling the uprising.

As an initial matter, we—like the magistrate judge and district court—have little trouble in agreeing that the alleged actions here occurred during the course of a violent political uprising. The Peruvian government and the Shining Path were engaged in a violent struggle for control of the country. According to one expert's opinion of the situation in 1992, "[a]pproximately 50 percent of Peruvian territory and approximately 65 percent of the country's population [was] under a state of national emergency." (J.A. at 108 (The Threat of the Shining Path to Democracy in Peru: Hearing Before the Subcomm. on Western Hemisphere Affairs of the H. Comm. on Foreign Affairs, 102d Cong. 17 (1992) (statement of Gordon H. McCormick, senior social scientist, RAND Corp.)).) Clearly, then, it is appropriate to describe the situation in Peru at the time of Ordinola's alleged actions as "a political revolt, an insurrection, or a civil war." Ornelas, 161 U.S. at 511, 16 S. Ct. 689.

The more difficult question is whether Ordinola's alleged offenses were incident to the political uprising. First, we recognize that it makes little sense to ask whether his actions were in furtherance of or "in aid" of the uprising. See id. Ordinola, acting on behalf of the Peruvian government, was attempting to defeat, not aid, the uprising. Accordingly, we must slightly alter the question and instead ask whether Ordinola's actions were incident to or in furtherance of quelling the violent uprising.

The parties disagree over the extent to which this test is subjective or objective. The district court, for example, focused in part on its independent factual finding that Ordinola had no "knowledge regarding the innocence of the civilians" at the time of the attacks and thus intended to quell the uprising via his actions. (J.A. at 526-27.) The Government contends that it is a mistake to focus on the intentions, or motives, of the accused. Instead, the Government argues, courts must concentrate on the act itself, as the Treaty exempts political offenses, not politically motivated offenders.

We conclude that courts must look at the question both subjectively and objectively, although the objective view must usually carry more weight. Courts have long recognized the relevancy of subjective motives in the political offense context. See, e.g., In re Castioni [1891] 1 Q.B. 149, 158 (1890) (Opinion of Denman, J.) ("It must at least be shown that the act is done in furtherance of, done with the intention of assistance, as a sort of overt act in the course of acting in a political matter, a political uprising, or a dispute between two parties in the State as to which is to have the government in its hands." (emphasis added)); Ornelas, 161 U.S. at 511, 16 S. Ct. 689 (inquiring whether the acts "were perpetrated with bona fide political or revolutionary designs"). We read these cases to mean that for a claimant to come within the protections of the political offense exception, it is necessary, but not sufficient, for the claimant to show that he was politically motivated. In other words, a claimant whose common crime was not subjectively politically motivated cannot come within the exception regardless of whether the offense itself could be described as an objectively "political" one.13

Aside from the subjective component, a claimant must also show that the offense was objectively political. See, e.g., Bassiouni, supra, at 602 (explaining that a test that focuses solely on motives "fails to appreciate the distinction between the nature of the offense and the motives of the actor"). This is because the Treaty itself exempts political offenses, and a political motivation does not turn every illegal action into a political offense. The Treaty, then, cannot be read to protect every act— no matter how unjustifiable and no matter the victim—simply because the suspect can proffer a political rationale for the action. See, e.g., Ahmad v. Wigen, 910 F.2d 1063, 1066 (2d Cir.1990) ("Political motivation does not convert every crime into a political offense."); Escobedo v. United States, 623 F.2d 1098, 1104 (5th Cir.1980) ("An offense is not of a political character simply because it was politically motivated.").

Here, we assume without deciding that Ordinola's actions were politically motivated. The magistrate judge did not explicitly address the question; instead, he focused on the objective side of the inquiry. The district court, on the other hand, appeared to base its grant of the writ on its independent conclusion that Ordinola was politically motivated. We see no reason to delve into this inquiry because, as explained below, we conclude that Ordinola is charged with offenses that are not political under the Treaty.

To determine whether a particular offense is political under the Treaty, we must look to the totality of the circumstances, focusing on such particulars as the mode of the attack and the identity of the victims, for example. In Ornelas, the Supreme Court examined Mexico's extradition request for a member of a band of armed men who attacked and killed a group of Mexican soldiers and civilians. Ornelas claimed that he was part of a border revolutionary movement and that his actions were in furtherance of that movement. Ornelas, 161 U.S. at 510-11, 16 S. Ct. 689. In a similar procedural posture to this case, the magistrate determined that the acts were not of a political character, but the district court disagreed on habeas review. In reviewing the Government's appeal, the Supreme Court queried whether the magistrate had "no choice, on the evidence, but to hold, in view of the character of the foray, the mode of attack, the persons killed or captured, and the kind of property taken or destroyed," that Ornelas's offenses were political. Id. at 511, 16 S. Ct. 689. In answering that question in the negative, the Supreme Court found it relevant that, inter alia, "private citizens" were the victims of some of the assaults, private property was taken, and Ornelas was part of a bandit group, acting without "uniforms or flag." Id. at 510, 16 S. Ct. 689.

In examining the Colina Group's modes of attack, the magistrate judge noted that Ordinola was being charged for his role in killing fifteen people when the Colina Group "opened fire on [a] crowd of men, women, and children." (J.A. at 19.) He also noted that Ordinola and other Colina Group agents made a group of students "dig trenches, forced them into the holes, killed them with gunfire, and buried them." (J.A. at 20-21.) Likewise, masked agents kidnapped a director of a news program, forced him to dig his grave, and killed him by shooting him in his head. Finally, the magistrate judge found that the agents acted on behalf of a friend of the military—who was perturbed by his workers' demands for higher wages—by kidnapping and murdering the men and then decorating the scene with Shining Path symbols to make it appear that the men had fallen victim to the Shining Path. The mode of these attacks does not favor Ordinola. They suggest a level of indiscriminate, clandestine killing based on suspect intelligence and political favors, as well as subsequent cover-ups that attempted to conceal the murders and destroy evidence. See, e.g., Matter of Extradition of Marzook, 924 F. Supp. 565, 577 (S.D.N.Y.1996) (rejecting political offense exception because of indiscriminate killings of civilians that approach "crimes abhorrent to human nature").

Ordinola fares no better when looking to "the persons killed." Ornelas, 161 U.S. at 511, 16 S. Ct. 689. None of the victims at issue here were armed at the time of attack or engaging in any overt hostility toward the Peruvian government. The magistrate judge accordingly found that "the victims in this case were clearly civilians, and of those who were alleged to have terrorist connections, evidence of such connections is tenuous at best." (J.A. at 31.) In so finding, the magistrate judge noted, inter alia, that the little boy who was killed while running to aid his dying father "was almost certainly not a terrorist, and the same can most likely be said for the other partygoers." (J.A. at 31.) The magistrate judge also found that the factory workers were targeted not because they were terrorists, but "because their boss wanted to retaliate against them for demanding better working conditions." (J.A. at 31.)

Recognizing the exceeding difficulty in attacking the sufficiency of the evidence supporting the magistrate judge's findings that the victims were not active members of the Shining Path, see Ornelas, 161 U.S. at 512, 16 S. Ct. 689 ("It is enough if it appear[s] that there was legal evidence on which the [magistrate judge] might properly conclude that the accused had committed offenses within the treaty as charged ...." (emphasis added)); cf. Anderson v. City of Bessemer, 470 U.S. 564, 574, 105 S. Ct. 1504, 84 L. Ed. 2d 518 (1985) (holding that a reviewing court may not reverse a plausible account of the evidence "even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently"), Ordinola instead contends that the civilian status of the victims is not especially relevant in this instance.14 We disagree.

Ordinola relies on language from Quinn v. Robinson,15 where the Ninth Circuit concluded that "there is no justification for distinguishing ... between attacks on military and civilian targets." Quinn, 783 F.2d at 810. In support of this "non-judgmental" contention, the court noted its view that "[i]t is for the revolutionaries, not the courts, to determine what tactics may help further their chances of bringing down or changing the government." Id.

We respectfully disagree with this conclusion and hold that there are in fact sound justifications for distinguishing between civilian and governmental—or in this case, revolutionary—targets. The first justification, of course, is that the Supreme Court has held that the civilian status of the "persons killed" is relevant. Ornelas, 161 U.S. at 511, 16 S. Ct. 689. Moreover, both the Second and Seventh Circuits have addressed the question and have concluded, like we do, that the status of the victims is relevant.16 See Ahmad, 910 F.2d at 1066 (holding that an attack on a commercial bus full of innocent civilians was not a political offense despite the existence of a political motivation); Eain, 641 F.2d at 521 (holding that "the indiscriminate bombing of a civilian populace is not recognized as a protected political act").

Second, it must be remembered that we are interpreting the Treaty to define the term "political offense." In doing so, we must afford "great weight" to the meaning attributed to the provision by the State Department, as it is charged with enforcing the Treaty. Sumitomo Shoji Am., Inc. v. Avagliano, 457 U.S. 176, 184-85, 102 S. Ct. 2374, 72 L. Ed. 2d 765 (1982); see also Al-Hamdi, 356 F.3d at 570 (granting "substantial deference" to the State Department's interpretation of provisions of the Vienna Convention). The State Department has previously expressed the view that "the political offense exception is not applicable to violent attacks on civilians," Ahmad v. Wigen, 726 F. Supp. 389, 402 (E.D.N.Y.1989), aff'd, 910 F.2d 1063 (2d Cir.1990), and at oral argument, Government counsel informed us that the department continues to generally adhere to that view.

Third, the status of the victims has been an important factor since the inception of the political offense exception. See, e.g., Ahmad, 726 F.Supp. at 404 ("In finding no distinction between targets, the Ninth Circuit ignored the fact that the civilian status of victims has been a significant factor in the political offense calculus since the nineteenth century."). In fact, aside from Ninth Circuit caselaw, we can find no other American authority stating that the civilian status of victims is irrelevant. Rather, the precedents and authorities stand firmly on the opposite front. See, e.g., Ornelas, 161 U.S. at 511, 16 S. Ct. 689; Ahmad, 910 F.2d at 1066; Eain, 641 F.2d at 520-21; Manuel R. Garcia-Mora, Crimes Against Humanity and the Principle of Nonextradition of Political Offenders, 62 Mich. L.Rev. 927, 944 (1964) ("[T]he means by which hostilities are conducted must be legitimately connected with the war to be characterized as political. It can scarcely be denied that the commission of atrocities during the conduct of hostilities or during belligerent occupation has no connection with furthering the legitimate policy of the State."); Steven Lubet & Morris Czackes, The Role of the American Judiciary in the Extradition of Political Terrorists, 71 J. Crim L. & Criminology 193, 202 (1980) ("[A]n offense having its impact upon the citizenry, but not directly upon the government, does not fall within the political offense exception.").

Fourth, by refusing to examine the scope of the attack, the mode of the attack, and the victims of the attack, the Ninth Circuit's approach results in defining a political offense as any common crime that occurs during a political uprising so long as the accused claims a political motive connected to the uprising. As explained above, we must reject such a subjective test. See also Eain, 641 F.2d at 521 (explaining the danger of abusing a treaty to protect criminals "simply because [their crimes] occurred during a time of political upheaval").

Accordingly, the magistrate judge's reasonable finding that Ordinola's alleged offenses were carried out against innocent civilians largely dooms Ordinola's argument. Because that finding was legitimately made, it simply cannot be said that the magistrate judge had no choice but to define Ordinola's alleged actions as political offenses.17 To have been considered political offenses, Ordinola's actions would had to have been in some way proportional to or in furtherance of quelling the Shining Path's rebellion. See id. at 521 (noting that the "legitimacy of a cause does not in itself legitimize the use of certain forms of violence especially against the innocent" (internal quotation marks and alteration omitted)). The magistrate judge did not err in recognizing that terror, for terror's sake, was not a sufficient method of quelling the Shining Path's uprising. Cf. id. ("The exception does not make ... the cold-blooded murder of civilians incidental to a purpose of toppling a government, absent a direct link between the perpetrator, a political organization's political goals, and the specific act."); Lubet & Czackes, supra, at 202 ("An offense which is intended only to disrupt the social order, but not to maintain or alter the government, is not political.").

Finally, both parties suggest that we make inferences based on the motives of the requesting government. For example, Ordinola contends that "a new government seeks to punish [him as a] member[ ] of a former government for [his] conduct in suppressing a violent uprising." (Appellant's Br. at 29.) The Government, for its part, points out that this is not a "new government" comprised of the revolutionaries Ordinola once fought; rather, it is the same democratically elected government—albeit a different administration— that requests Ordinola's extradition. Although the Government's interpretation is correct on the facts, the motives of the requesting government are irrelevant to our decision. The Treaty states that extradition will be denied "if the executive authority of the Requested State determines that the request was politically motivated." Extradition Treaty, art. IV, sec. 3 (emphasis added). Any question into the Peruvian government's motivations is therefore well beyond this Court's legitimate realm of authority under the Treaty and must be addressed solely to the Secretary of State.18 See Eain, 641 F.2d at 513 ("It is the settled rule that it is within the Secretary of State's sole discretion to determine whether or not a country's requisition for extradition is made with a view to try or punish the fugitive for a political crime, i.e., whether the request is a subterfuge."); cf. United States v. Kin-Hong, 110 F.3d 103, 111 (1st Cir.1997) ("[I]t is not that questions about what awaits the relator in the requesting country are irrelevant to extradition; it is that there is another branch of government, which has both final say and greater discretion in these proceedings, to whom these questions are more properly addressed.").

For the reasons outlined above, the district court erred in granting Ordinola's petition for a writ of habeas corpus. In affording proper deference under habeas review to the magistrate's decision, we simply cannot say that the magistrate judge had "no choice" but to find that the common crimes Ordinola is alleged to have committed against noncombatant civilians were political offenses under the Treaty with Peru. We therefore must vacate the district court's grant of the writ, and remand for reentry of a Certification of Extraditability. We note, however, that judicial review is not the only recourse available to Ordinola, as he is free to contest his extradition before the United States Secretary of State.

VACATED AND REMANDED.

I agree that the magistrate judge correctly certified pursuant to 18 U.S.C.A. § 3184 (West Supp.2006) that Ordinola may be surrendered to Peru under the terms of the Extradition Treaty between the United States and Peru. Accordingly, I concur in the result reached by the majority opinion vacating the grant of habeas relief by the district court and directing reentry of the certification of extraditability. Because I would follow a different analytical approach, however, I write separately.

The Judiciary's Limited Role in International Extradition Cases

Extradition is "the surrender by one nation to another of an individual accused or convicted of an offense outside of its own territory, and within the territorial jurisdiction of the other [nation], which ... demands the surrender." Terlinden v. Ames, 184 U.S. 270, 289, 22 S. Ct. 484, 46 L. Ed. 534 (1902). There is no general duty under international law, however, for a sovereign nation to surrender an accused fugitive upon demand. See Factor v. Laubenheimer, 290 U.S. 276, 287, 54 S. Ct. 191, 78 L. Ed. 315 (1933). The right of one sovereign to demand that another sovereign extradite an accused criminal arises, if at all, by treaty; thus, no duty to surrender an alleged fugitive to a foreign government exists apart from an extradition treaty. See id.; see also United States v. Alvarez-Machain, 504 U.S. 655, 664, 112 S. Ct. 2188, 119 L. Ed. 2d 441 (1992) ("Extradition treaties exist so as to impose mutual obligations to surrender individuals in certain defined sets of circumstances, following established procedures.").

Because extradition is a creature of treaty, "the power to extradite derives from the President's power to conduct foreign affairs." Sidali v. INS, 107 F.3d 191, 194 (3d Cir.1997); see U.S. Const. art. II, § 2, cl. 2. Extradition, therefore, is an executive function rather than a judicial one. See, e.g., Martin v. Warden, 993 F.2d 824, 828 (11th Cir.1993). It involves a "diplomatic process carried out through the powers of the executive, not the judicial, branch." Blaxland v. Commonwealth Dir. of Pub. Prosecutions, 323 F.3d 1198, 1207 (9th Cir.2003). The decision to extradite is one that is "entirely within the discretion of the executive branch, except to the extent that the statute interposes a judicial function." Lopez-Smith v. Hood, 121 F.3d 1322, 1326 (9th Cir.1997).

Still, the judiciary plays a limited role in the overall extradition process, as prescribed by Congress in the Extradition Act. See 18 U.S.C.A. §§ 3181-3186 (West 2000 & Supp.2006); see Sidali, 107 F.3d at 194 ("[T]he judiciary has no greater role than that mandated by the Constitution, or granted to the judiciary by Congress." (internal quotation marks omitted)). The essential function served by an extradition court under the statute is "to determin[e] an individual's eligibility to be extradited," meaning that the court must "ascertain[] whether a crime is extraditable under the relevant treaty and whether probable cause exists to sustain the charge." Vo v. Benov, 447 F.3d 1235, 1245 (9th Cir.), cert. denied, ___ U.S. ____, 127 S. Ct. 317, 166 L. Ed. 2d 238 (2006). A primary purpose of the extradition statute is to "interpos[e] the judiciary between the executive and the individual," Austin v. Healey, 5 F.3d 598, 604 (2d Cir.1993), as well as provide a means of "independent review" of executive action. Martin, 993 F.2d at 828. The extradition court's statutory inquiry is simply a "preliminary examination to determine whether a case is made out which will justify the holding of the accused and his surrender to the demanding nation." Lo Duca v. United States, 93 F.3d 1100, 1104 (2d Cir.1996) (internal quotation marks omitted). Therefore, the "judicial officer who conducts an extradition hearing ... performs an assignment in line with his or her accustomed task of determining if there is probable cause to hold a defendant to answer for the commission of an offense." Id. (internal quotation marks omitted); cf. In re McMullen, 989 F.2d 603, 611 (2d Cir.1993) (en banc) ("What is at issue in the proceeding ... is not punishability but prosecutability." (internal quotation marks omitted)). Thus, there is no direct appeal from the extradition court's determination that an individual is or is not subject to extradition habeas corpus provides the only means of review. See Eain v. Wilkes, 641 F.2d 504, 508-09 (7th Cir.1981).

Ultimately, the decision whether to surrender a person found eligible for extradition remains a discretionary one committed to the executive branch. The judiciary has no power to order the extradition of the alleged fugitive. After the extradition court completes its limited inquiry, the executive, through the Secretary of State, makes a discretionary decision whether extradition, although permissible under the statute, is appropriate based on "factors affecting both the individual defendant as well as foreign relations in deciding whether extradition is appropriate." Hoxha v. Levi, 465 F.3d 554, 563 (3d Cir.2006) (internal quotation marks omitted).

In addition to the limits established by statute, federal courts have developed various self-limiting principles that "ensure, among other things, that the judicial inquiry does not unnecessarily impinge upon executive prerogative and expertise." United States v. Kin-Hong, 110 F.3d 103, 110 (1st Cir.1997). These judicially-created doctrines keep extradition courts within the bounds of the limited judicial inquiry set by statute and serve as a barrier to the expansion of the judiciary's role through habeas review.

One such limiting doctrine, for example, is the rule of non-inquiry, pursuant to which courts refrain from delving into and assessing the competence of the requesting government's system of justice. See id. at 110-11. Questions about the procedural fairness of another sovereign's justice system or whether the individual to-be-surrendered faces inhumane treatment are within the purview of the executive branch. See id. at 111 ("It is not that questions about what awaits the relator in the requesting country are irrelevant to extradition; it is that there is another branch of government, which has both final say and greater discretion in these proceedings, to whom these questions are more properly addressed."). Likewise, it is a question for the executive branch, not the courts, whether the requesting nation is sincere in its demand for extradition or is merely using the process as a subterfuge to exact revenge against an opponent of the government. See Eain, 641 F.2d at 513. The rule of non-inquiry, then, "serves interests of international comity by relegating to political actors the sensitive foreign policy judgments that are often involved in the question of whether to refuse an extradition request." Hoxha, 465 F.3d at 563.

Courts also assume a deferential posture when it comes to determining the existence or continuing validity of an extradition treaty on the grounds that such questions are essentially political. See id. at 562. Courts accord the executive branch's construction of a treaty "great weight," Kin-Hong, 110 F.3d at 110, largely because the executive branch wrote and negotiated the document being interpreted, see In re Howard, 996 F.2d 1320, 1330 n. 6 (1st Cir.1993).

In sum, federal courts—whether it be a magistrate judge serving as the extradition court or a district judge sitting in habeas—are tasked with resolving a single issue—does the purported fugitive appear to be eligible for extradition under the relevant treaty?

Extradition Proceedings under 18 U.S.C. § 3184

Extradition in the United States is generally initiated by the submission of an extradition request from a foreign government to the United States Department of State, pursuant to the relevant treaty. See Barapind v. Reno, 225 F.3d 1100, 1105 (9th Cir.2000). This request must generally "be supported by sufficient evidence to show that the individual is the person sought for the crimes charged, that the crimes are among those listed as extraditable offenses in the Treaty and that there is sufficient justification for the individual's arrest had the charged crime been committed in the United States." Eain, 641 F.2d at 508. The State Department makes an initial evaluation to determine whether the request falls within the scope of the treaty, see Vo, 447 F.3d at 1237, and then turns the matter over to the appropriate United States Attorney for the filing of a complaint seeking a certificate of extradition under 18 U.S.C.A. § 3184, see Eain, 641 F.2d at 508.

From that point, the U.S. Attorney essentially represents the foreign government seeking return of the alleged fugitive in all extradition litigation. The statute directs the U.S. Attorney to submit the complaint for an arrest warrant "under oath" to an extradition officer, i.e., "any justice or judge of the United States, or any magistrate authorized ... by a court of the United States." 18 U.S.C.A. § 3184. In practice, this is generally a magistrate judge who holds a limited "hearing to determine whether (1) the crime is extraditable; and (2) there is probable cause to sustain the charge." Prasoprat v. Benov, 421 F.3d 1009, 1012 (9th Cir.2005) (internal quotation marks omitted), cert. denied, ___ U.S. ____, 126 S. Ct. 1335, 164 L. Ed. 2d 51 (2006); see 18 U.S.C.A. § 3184. The extradition hearing, of course, "`is not ... in the nature of a final trial by which the prisoner could be convicted or acquitted of the crime charged against him.'" LoDuca, 93 F.3d at 1104 (quoting Benson v. McMahon, 127 U.S. 457, 463, 8 S. Ct. 1240, 32 L. Ed. 234 (1888)). Stated differently, the hearing is "not designed as a full trial" but as a means of "inquir[ing] into the presence of probable cause to believe that there has been a violation of one or more of the criminal laws of the extraditing country." Peroff v. Hylton, 542 F.2d 1247, 1249 (4th Cir.1976); see Eain, 641 F.2d at 508 ("It is fundamental that the person whose extradition is sought is not entitled to a full trial at the magistrate's probable cause hearing . . . . That is the task of the . . . courts of the other country.").

Although the extradition statute does not mention "probable cause" and instead directs the extradition court to determine whether there is "evidence sufficient to sustain the charge under the provisions of the proper treaty or convention," 18 U.S.C.A. § 3184, courts have uniformly interpreted the statutory language to require a finding of "probable cause." See Vo, 447 F.3d at 1237; Sidali, 107 F.3d at 195. Thus, "[t]he probable cause standard applicable to an extradition hearing is the same as the standard used in federal preliminary hearings," meaning that the magistrate judge's role is merely "to determine whether there is competent evidence to justify holding the accused to await trial." Hoxha, 465 F.3d at 561 (internal quotation marks omitted). In that vein, the evidence considered by the magistrate as part of an extradition hearing "need not meet the standards for admissibility at trial" and "may be based upon hearsay in whole or in part." Kin-Hong, 110 F.3d at 120 (internal quotation marks omitted).

Not only are the admissibility standards relaxed, but the alleged fugitive's ability to challenge the government's evidence or to submit evidence of his own at the extradition hearing is also significantly limited. For example, the fugitive has no right to cross-examine witnesses, see Oen Yin-Choy v. Robinson, 858 F.2d 1400, 1406-07 (9th Cir.1988), or to introduce "contradictory evidence" that conflicts with the government's probable cause evidence, see Hoxha, 465 F.3d at 561. By contrast, "explanatory evidence" relating to the underlying charges is admissible. See Koskotas v. Roche, 931 F.2d 169, 175 (1st Cir.1991).1

If the extradition judge concludes that there is, in fact, probable cause, he "is required to certify the individual as extraditable to the Secretary of State." Vo, 447 F.3d at 1237 (internal quotation marks omitted). As previously noted, once the judicial officer determines that an individual is eligible for extradition, the matter returns to the Secretary of State for the executive branch to make the ultimate decision whether to surrender the requested fugitive. See, e.g., Cheung v. United States, 213 F.3d 82, 88 (2d Cir.2000); see 18 U.S.C.A. § 3186 (West 2000) ("The Secretary of State may order the person committed under section 3184 ... to be delivered to any authorized agent of [the requesting] foreign government ...." (emphasis added)).

Finally, as noted previously, part of the extradition court's duty is to determine whether the crime charged is covered by the particular extradition treaty as an extraditable offense. Naturally, this determination may require the extradition judge to decide whether the charged offense falls within the political offense exception. If it does, then the offense is a non-extraditable one and the accused fugitive cannot be eligible for extradition, despite the existence of probable cause. See, e.g., Vo, 447 F.3d at 1237-38. Such a decision effectively removes discretion from the Secretary of State, who may not extradite in the absence of a certificate of extraditability. See 18 U.S.C.A. § 3186. The government, however, may continue to pursue the necessary certification by presenting its case to a different magistrate judge. See Collins v. Loisel, 262 U.S. 426, 430, 43 S. Ct. 618, 67 L. Ed. 1062 (1923).

Habeas Review of the Certification of Extraditability

Assuming the extradition judge certifies the extradition request to the Secretary of State, the fugitive has but one means of judicial recourse—filing a habeas petition under 28 U.S.C.A. § 2241 (West 2000). See Collins v. Miller, 252 U.S. 364, 369-70, 40 S. Ct. 347, 64 L. Ed. 616 (1920); see also Kastnerova v. United States, 365 F.3d 980, 984 n. 4 (11th Cir.2004) ("There is no direct appeal in extradition proceedings."); Plaster v. United States, 720 F.2d 340, 347-48 (4th Cir.1983).

Habeas review in the extradition context is very narrow. See Kastnerova, 365 F.3d at 984. Essentially, habeas review of the extradition court's certification is limited to consideration of whether the extradition court had jurisdiction, whether the charged crime qualifies as an extraditable offense under the treaty, and whether there is "any evidence" to support the magistrate judge's finding of probable cause. Prushinowski v. Samples, 734 F.2d 1016, 1018 (4th Cir.1984) (quoting Fernandez v. Phillips, 268 U.S. 311, 312, 45 S. Ct. 541, 69 L. Ed. 970 (1925)); see Kastnerova, 365 F.3d at 985 (explaining that the habeas court's job is to determine "whether the Government presented competent evidence upon which the magistrate could find there were reasonable grounds upon which to believe [the accused fugitive] guilty of the charged offenses"). Thus, the standard of review applied to the decision of a magistrate judge to issue a certificate of extradition is at least as deferential, if not more so, than that applied to a magistrate judge's decision to issue a search warrant. See Illinois v. Gates, 462 U.S. 213, 236, 103 S. Ct. 2317, 76 L. Ed. 2d 527 (1983) (requiring appellate courts to affirm the issuance of a warrant if there is a "substantial basis" to support the magistrate judge's probable cause determination (internal quotation marks omitted)). It is well-established that a reviewing court is not permitted to substitute its own view of the facts for the findings of the magistrate judge. See United States v. Fuller, 441 F.2d 755, 759 (4th Cir.1971).

In no sense does a habeas action for the review of an extradition decision afford the reviewing court an opportunity to weigh the evidence or serve in a fact-finding capacity. Just as the magistrate judge's underlying determination is not a mini-trial on the guilt or innocence of the fugitive, the district court's habeas review should not duplicate the extradition hearing—habeas review "is not a means for rehearing the magistrate's findings." Afanasjev v. Hurlburt, 418 F.3d 1159, 1163 (11th Cir.) (internal quotation marks omitted), cert. denied, ___ U.S. ____, 126 S. Ct. 587, 163 L. Ed. 2d 489 (2005). As long as the factual findings that support the magistrate judge's legal conclusions are not clearly erroneous, the reviewing court is bound by the facts as determined by the magistrate. The reviewing court may not substitute its own assessment of the facts for that of the magistrate judge nor may it make additional, new findings of fact that the extradition court did not make.

In this case, the only real question is whether the charged crime is an extraditable offense under the treaty—if Ordinola is charged with a political offense, it is not an extraditable crime. The magistrate judge's determination in this regard required factual findings and the application of law to such facts; therefore, we are presented with a mixed question of law and fact. See Ornelas v. Ruiz, 161 U.S. 502, 509, 16 S. Ct. 689, 40 L. Ed. 787 (1896); see also Vo, 447 F.3d at 1240. We review the extradition court's factual findings for clear error and its conclusions of law de novo. See Afanasjev, 418 F.3d at 1163.

Under the bilateral extradition treaty, the United States and Peru "agree to extradite to each other ... persons whom the authorities in the Requesting State have charged with, found guilty of, or sentenced for, the commission of an extraditable offense." Extradition Treaty with Peru, U.S.-Peru, July 26, 2001, S. Treaty Doc. 107-6 (2002) [hereinafter Peru Treaty], art. I. The treaty defines an extraditable offense as an offense that is "punishable under the laws in both Contracting States by deprivation of liberty for . . . more than one year or by a more severe penalty," Peru Treaty, art. II, § 1, or the attempt to commit such an offense, see Peru Treaty, art. II, § 2.

Article IV of the extradition treaty also sets forth several specific bases for denying an extradition request, including the mandatory denial of extradition for political offenses: "Extradition shall not be granted if the offense for which extradition is requested constitutes a political offense." Peru Treaty, art. IV, § 2. As is typical of international extradition treaties, however, the bilateral extradition agreement between the United States and Peru does not define what constitutes a "political offense." Instead, it specifies three offenses that do not qualify as a political offense: (1) murder "of a Head of State of one of the Contracting States"; (2) "genocide, as described in the Convention on the Prevention and Punishment of the Crime of Genocide"; and (3) "an offense for which both Contracting States have the obligation pursuant to a multilateral international agreement to extradite the person sought or to [prosecute] the case...." Id.2

It has fallen to the courts to develop and establish the contours of the political offense exception to the duty to extradite. See Barbara Sicalides, Comment, RICO, CCE, and International Extradition, 62 Temp. L.Rev. 1281, 1301 (1989) ("Because there is no international agreement concerning which crimes constitute political offenses and because the term is seldom defined in treaties, its definition and application have been left to the courts."); M. Bassiouni, International Extradition and World Public Order 322, 371 (1974) (noting that the judiciary is the primary source for developing the parameters of the political offense exception). As the majority opinion ably explains, federal courts recognize a distinction between "pure" and "relative" political offenses. See Quinn v. Robinson, 783 F.2d 776, 793 (9th Cir.1986). Ordinola argues that he is charged with crimes that would qualify as relative political offenses, i.e., common crimes that are "so connected with a political act that the entire offense is regarded as political." Eain, 641 F.2d at 512 (internal quotation marks omitted). "Pure" political offenses such as espionage or treason are rarely, if ever, at issue in extradition litigation; it is the "relative" category involving otherwise extraditable common crimes that produces significant "definitional problems." Quinn, 783 F.2d at 793-94. The question of whether a common crime is sufficiently connected to a political act would seem to require courts to venture beyond their limits in the extradition process and into the political domain of the executive branch. See Eain, 641 F.2d at 513 (noting the anomalous nature of the judicial branch's historical practice of deciding whether a crime charged under the law of a foreign sovereign is a political offense). Nevertheless, the long-established practice of the judiciary is to determine whether the political offense exception applies to a given common crime. See Ornelas, 161 U.S. at 502, 16 S. Ct. 689.

The primary test for determining whether a crime is a relative political offense, as noted by the majority, involves two questions: (1) whether, at the time of the alleged offense, there was a "violent political disturbance, such as war, revolution and rebellion"; and (2) whether the alleged offense was "committed in the course of and incidental to" the violent political disturbance. Escobedo v. United States, 623 F.2d 1098, 1104 (5th Cir.1980). Virtually every court to encounter the question of whether the alleged crime is a relative political offense has applied this two-pronged test. See, e.g., Barapind v. Enomoto, 400 F.3d 744, 750 (9th Cir.2005) (en banc).

The political offense exception provides the would-be extraditee with an affirmative defense to extradition. See Vo, 447 F.3d at 1242. Thus, once the government carries its burden of establishing that the defendant is subject to extradition, the burden shifts to the extraditee to prove that the crimes with which he is charged are political offenses. See id.

As thoroughly detailed by Judge Williams, Ordinola was charged in Peru with Aggravated Homicide and Aggravated Kidnapping, among other offenses, arising out of Ordinola's association with the Colina Group, a counter terrorist unit within the Peruvian Army assigned to combat the Shining Path, a violent revolutionary group bent on destroying Peru's government. The charges against Ordinola stem from four underlying incidents that occurred in 1991 and 1992 in which Ordinola and other members of the Colinas allegedly brutalized and murdered numerous individuals.

Ordinola claims that these offenses cannot serve as the bases for extradition because he committed the underlying acts in the course of his duty as a Colina Group member to fight the terrorist threat posed by the Shining Path. Ordinola suggests that he was merely carrying out his duty to preserve the government in the midst of political upheaval; therefore, he contends that the underlying crimes with which he is charged qualify as "political offenses" within the meaning of the Extradition Treaty between the United States and Peru.

Ordinola's defense to extradition raises an interesting theoretical issue about the application of the political offense exception to former government agents charged with crimes committed while there is an ongoing political uprising against the government. See Aimee J. Buckland, Comment, Offending Officials: Former Government Actors and the Political Offense Exception to Extradition, 94 Cal. L.Rev. 423, 424 (2006) ("[T]he law is unsettled as to whether former government officials may use the political offense exception and whether the exception is appropriate in this context."). "The original purpose of the political offense exception to extradition was to protect revolutionaries from being returned to . . . face prosecution for crimes committed against their governments," id. at 423, not to afford protection to those seeking to suppress a rebellion. There is support for the position that the political offense exception does not apply in a situation such as Ordinola's because, "[a]rguably, everything a government actor does is `political' ... [and therefore] the exception theoretically protects former government officials from facing their accusers merely because they used to govern them." Id. at 424. See In re Extradition of Suarez-Mason, 694 F. Supp. 676, 705 (N.D.Cal.1988) (concluding that "the principles underlying the political offense exception are ill-served by extending its protection to former government officials").

Unfortunately, the issue of whether the political offense exception should ever be applied to a former government actor is not technically before this court, the government having elected not to pursue the argument before the extradition court and on habeas review. Accordingly, we have no occasion to address this difficult problem directly and must assume the exception is theoretically available to an extraditee such as Ordinola. As a practical matter, however, the difficulty of assessing a former government agent's claim to the political offense exception persists in Ordinola's argument, which is essentially no different than the claim that any acts he committed under the auspices of the Colina Group were political. Indeed, the sole purpose for the existence of the Colina Group was political: to preserve the government from the revolutionary group that sought to overthrow it. Specifically, Ordinola takes the position that he is charged with political offenses because the underlying acts were politically motivated—he was acting pursuant to directives from his superior officers. Thus, in Ordinola's mind, he was fighting an insurgency. Ordinola's argument seeks to reduce the test for determining whether the charged crime falls within the political offense exception to one based on subjective motivation. But, if we are going to permit a former government actor to avail himself of the political offense exception, the offender's subjective motivation cannot serve as the sole determinate of whether an offense is a political one. As Judge Williams correctly noted, "[a]n offense is not of a political character simply because it was politically motivated." Escobedo, 623 F.2d at 1104; see Ahmad v. Wigen, 910 F.2d 1063, 1066 (2d Cir.1999); Eain, 641 F.2d at 520 ("[F]or purposes of extradition, motivation is not itself determinative of the political character of any given act.").

Rather, the heart of the inquiry when it comes to determining whether the charged offense falls within the political offense exception must be objective. See Ornelas, 161 U.S. at 509-12, 16 S. Ct. 689. Factors such as "the character of the foray, the mode of attack, [and] the persons killed or captured" are appropriate for courts to consider in determining whether, as an objective matter, the common crime at issue was "incidental" to a major political disturbance or uprising. Id. at 511-12, 16 S. Ct. 689.

The magistrate judge concluded that there was a political uprising sufficient to meet the first prong of the incidence test. He concluded further, however, that none of the four underlying events were sufficiently incidental to the uprising to qualify as political offenses. The magistrate judge's conclusion that Ordinola's charged offenses were not "political offenses" rested primarily on his conclusion that the "victims in this case were clearly civilians" with no connection to the political upheaval in Peru at the time. J.A. 31. Sitting in review of the extradition order, the district court concluded to the contrary—that Ordinola was charged with a political offense because he "did not knowingly murder innocent civilians" but was responding to the threat posed by the Shining Path, J.A. 528.3 Although the district court's assessment of the evidence presented to the magistrate judge as part of the extradition hearing is certainly plausible, it ignores the question of whether the magistrate judge's assessment was clearly erroneous. Rather than reviewing the extradition court's factual findings for clear error, the district court ruled based on its preferred view of the record—that Ordinola's targets were not necessarily civilians. Additionally, the district court relied upon a new factual determination—that Ordinola did not act knowingly. The magistrate judge did not make an explicit finding regarding Ordinola's knowledge of the civilian status of his victims. If anything, the magistrate judge implicitly declined to make such a finding, noting that subjective "political intentions on the part of the extraditee are not enough to render an act a political offense." J.A. 29. Moreover, the district court's reliance on this later finding of fact—Ordinola's lack of knowledge—rests on the same faulty premise as Ordinola's argument that his subjective motivation is a determining factor.

In any event, our review of the extradition decision must consider the factual basis as determined by the magistrate judge without regard to the district court's view of the facts.

In May 1991, Ordinola and other members of the Colina Group allegedly killed 15 people, including women and children, at a fundraiser purportedly being held to benefit the Shining Path. The magistrate judge found that the Colinas, riding in two vehicles, pulled up at an apartment complex in the Barrios Altos section in Lima where the party was proceeding in the complex's yard and "opened fire on the crowd of men, women, and children." J.A. 19. Ordinola was identified by another Colinas member as a participant in the attacks and specifically as the shooter of a child who was running to his wounded father. The magistrate judge concluded that "the little boy whom [Ordinola] allegedly shot at the [fundraiser] ... was almost certainly not a terrorist, and the same can most likely be said for the other partygoers." J.A. 31.

Ordinola does not contradict these facts. Rather, he contends that other evidence submitted by the Peruvian government demonstrates that the Colina Group was acting based on information provided by a government agent who infiltrated the Shining Path and reported that members and friends of the Shining Path gathered socially where the Barrios Altos killings occurred. Ordinola also highlights Peru's acknowledgment, in its extradition submission to the State Department, that it "presumed that the murder and bodily injuries caused ... were motivated by a feeling of retaliation against terrorism." J.A. 441. Ordinola suggests, therefore, that his actions as a member of the Colinas were undertaken in response to specific threats posed by the Shining Path in order to protect the Peruvian government. Thus, he argues that his crimes were incidental to the political disturbance in Peru at the time.

The magistrate judge rejected Ordinola's position, concluding that "mere political intentions on the part of the extraditee are not enough to render an act a political offense." J.A. 29. In determining that Ordinola's crimes were not incidental to the ongoing political disturbance in Peru, the magistrate judge listed the objective Ornelas factors but discussed only one of them—the status of the persons killed or captured. The magistrate judge relied heavily on his determination that the victims were civilians without any real connections to the Shining Path.

I would agree that Ordinola failed to demonstrate that these actions were "incidental" to a political disturbance. His argument is essentially that his conduct with respect to the Barrios Altos slayings was incidental to Peru's political upheaval because his actions were intended to further the Colinas' overarching objective of destroying the Shining Path. As suggested by the magistrate judge and explained in the majority opinion, Ordinola's subjective motivation, political though it may have been, does not alone convert his common crimes into political offenses. See Ahmad, 910 F.2d at 1066. Ordinola must show that his crimes were incidental to the political uprising from an objective vantage point, as well. He fails to do so. In fact, Ordinola does not really challenge the magistrate judge's finding of fact that the victims were civilians with no ties to the Shining Path, and he has failed to demonstrate that other Ornelas factors, such as the nature of the attack, suggest a political crime. Actually, the facts noted by the magistrate judge suggest precisely the opposite—the attack occurred in a residential area being used for a social gathering and was not connected to military operations against the government.

In May 1992, Peruvian Army General Juan Rivero arranged a meeting between Colina Group leaders and Jorge Fung Pineda, the owner of a private cotton mill. Cotton mill workers were making demands for higher wages and better machinery. Pineda, who had connections in the military, requested that the Colina Group, for a fee, "get [the workers] involved in terrorism and teach them a good lesson." J.A. 23. Members of the Colina Group, including Ordinola, kidnapped eight workers from their homes and a ninth while he was riding his bicycle, executed them and buried the bodies. At the burial location, the Colinas used red paint to write Shining Path slogans.

Ordinola argues that he and the other Colinas acted under the belief, based on information provided by government collaborators, that the mill workers were involved in "subversive" activities. He suggests that the evidence shows the Colinas were kept in the dark about the true nature of their mission by superior officers. According to a government witness, the Colinas believed they were executing terrorists and learned that the executions had been a "private job" only after the fact.

Again, Ordinola's argument hinges solely upon his subjective belief that he was fighting the government's war on terrorism, and that the cotton mill workers were actually insurrectionists. But, such facts do not demonstrate that the magistrate judge clearly erred in finding that Ordinola's actions were not incidental to a political uprising. That Ordinola was acting upon orders from superior officers or mistakenly believed that he was killing terrorists may ultimately demonstrate that he lacked the intent to commit murder or kidnapping; however, these issues are ones for the Peruvian justice system to sort out after a full-blown trial. For our part, we are not concerned with whether Ordinola has a defense to the underlying charges. Our job is simply to decide whether Ordinola is charged with a political offense based on the limited evidence adduced at the extradition hearing, which is preliminary in nature and not designed to fully develop the evidence.

In July 1992, Ordinola and other members of the Colina Group went to the dormitories at "La Cantutu" University, where an army agent had allegedly infiltrated and reported the presence of terrorists. Wearing masks and carrying weapons, shovels and lime, Ordinola and other Colinas removed fifty students from their rooms and assembled them outside. The army agent identified ten of the fifty students as terrorists. The Colinas also abducted a professor identified by the agent as a terrorist. The students and the professor were taken to another location, where they were executed after digging their own graves. Members of the Colina Group subsequently exhumed and incinerated the bodies for fear that they would be discovered.

In the final incident, which occurred in June 1992, Army Major Rivas summoned the Colina Group, including Ordinola, and informed them that they would conduct an operation in Huacho but did not provide any specifics. The following day, the Colina Group traveled in two vans to an area near a beach. Rivas directed one group to dig a grave and remain at the beach; the other group was ordered to abduct Pedro Yauri Bustamante from his home and return with him to the beach. Bustamante was a journalist who broadcast a daily radio program in which he often criticized the government. Army Intelligence had investigated Bustamante on two previous occasions and listed him as active terrorist.

The magistrate judge found that, although "it is slightly more plausible that the university students and professor and the journalist . . . had ties to the Shining Path [than did fundraiser attendees or the factory workers], the evidence [of such ties] is not strong." J.A. 31.

Ordinola's response, again, is that the Colina Group acted according to what it was told—that the targeted individuals were part of the Shining Path. Because the mission of the Colina Group was to eliminate the threat posed by the Shining Path, Ordinola argues that these incidents were necessarily connected to the political uprising. For the very same reasons set forth previously, I would reject this argument as well. Ordinola relies on his subjective motivation alone to prove that his crimes were political offenses; he fails to offer objective factors that demonstrate the political nature of the charged offenses.

In conclusion, when the magistrate judge viewed the evidence objectively, he found it insufficient to show that the killings and kidnappings were to suppress the Shining Path revolution. Men, women, and children had been killed indiscriminately. Shining Path slogans had been painted on the walls after the fact. Bodies were hidden or incinerated. Not one shred of objective evidence showed any of the victims to have been members or supporters of the Shining Path. Evidence that Ordinola was told by informants or superiors that the persons to be killed were insurgents or their supporters is not objective proof that they were. Instead, that evidence goes to what Ordinola subjectively believed and may be of relevance by way of defense to the charges if he is ultimately returned to Peru for trial. Here, however, it cannot avail him. Consequently, I can find no clear error in the conclusion of the magistrate judge that these crimes were not political offenses within the meaning of the extradition treaty. Therefore, I agree reversal of the district court is warranted.