dissenting:

The question in this case is, in the immortal words of Humpty Dumpty, which is to be the master — that’s all. When it comes to the granting of asylum, Congress has said the BIA is the master. The *1002 statute provides it, the other courts of appeals recognize it and the Supreme Court keeps reminding us of it. But to no avail. Maybe there’s something in the water out here, but our court seems bent on denying the BIA the deference a reviewing court owes an administrative agency. Instead, my colleagues prefer to tinker — to do the job of the Immigration Judge and the BIA, rather than their own. See, e.g., INS v. Chen, 537 U.S. 1016 , 123 S.Ct. 549 , 154 L.Ed.2d 423 (2002); INS v. Ventura, 537 U.S. 12 , 123 S.Ct. 353 , 154 L.Ed.2d 272 (2002); INS v. Elias-Zacarias, 502 U.S. 478 , 112 S.Ct. 812 , 117 L.Ed.2d 38 (1992).

This is yet another case in point. After carefully considering the record and correctly identifying the applicable law, the IJ concluded that petitioner is not eligible for asylum; the BIA affirmed, based largely on the IJ’s written opinion. The agency did not misunderstand the law or overlook key evidence; the agency did everything just right. What my colleagues find fault with, rather, is the process at the very heart of the agency’s authority and expertise: determining the basic facts under-girding an applicant’s asylum claim.

The IJ and the BIA made two findings that fatally undermine petitioner’s asylum claim: first, that whatever harassment petitioner may have suffered in his native country was not on account of his political opinion; and, second, that the harassment was not by the government. Not so fast, say my colleagues. We know such things much better than the IJ and the BIA, so we’re going to find the facts ourselves. But, isn’t this what we got slammed for just last Term? Well, never mind. The government can’t bother the Supremes every time we go over the top, so it’s a fair bet that if we keep marching to our own drummer we’ll mostly get away with it. Being the circuit with more asylum cases than all others combined, see INS v. Chen, Pet. for Cert. 2002 WL 32101070 at *29, available at http://www.usdoj.gov/osg/ briefs/2002/2pet/7pet/2002-0025.pet.aa.pdf, gives us molto institutional leverage.

But it’s not right. We have our job to do, and the BIA and the IJs have theirs. The process cannot work as Congress intended if we keep usurping the agency’s job rather than doing our own. This is a simple case with simple issues, most controlled by precedent; due deference to the agency’s functions calls for a straightforward affirmance of the agency’s reasonable decision.

1. Petitioner claims that, for a brief period in his youth, he was affiliated with a group called the Mojahedeen. The IJ found that petitioner “was not an official member. Rather, he belonged to a group that supported the Mojahedeen, distributed literature for them, sold newspapers and attended meetings.” A.R. at 160. 1 He quit in 1982, got a university degree and worked for a number of years for the Iranian government. Then, in 1990, he was approached by a soldier in the Iranian Revolutionary Guard, an outfit also known as the Pastars. This soldier recognized him as a former Mojahedeen affiliate and attempted to blackmail him with this information. The soldier threatened that, if petitioner did not pay a large sum of money, the soldier would report him to the government and petitioner would be arrested. The question presented is *1003 whether this encounter must be considered persecution for purposes of our asylum statute.

Reprehensible though the soldier’s conduct was, it does not amount to persecution under the asylum statute unless it satisfies two statutory criteria: it must have been “on account of ... political opinion,” 2 8 U.S.C. § 1101 (a)(42)(A), and it must have been done on behalf of the government, Sangha v. INS, 103 F.3d 1482, 1487 (9th Cir.1997). 3

The IJ carefully considered the evidence and ruled against petitioner on both grounds. She found the soldier did not blackmail petitioner on account of his political opinion, but was “motivated by his isolated desire for money.” A.R. at 165. The IJ also found the soldier was not “acting in concert with the government,” but was engaged in a private act of extortion — an attempt to get petitioner to pay money that the soldier would use entirely for his private purposes. Id. at 166. Any court applying the proper standard of review and giving the administrative agency the deference to which it is entitled would readily accept both of these findings.

a. Whether persecution is “on account of’ a petitioner’s political opinion is a question of fact; it turns on evidence about the persecutor’s motives. Here, the IJ found that “[t]he actions of the soldier appeared motivated by his isolated desire for money, not by the applicant’s political opinion. Furthermore, the applicant’s failure or refusal to pay the bribes was not politically motivated.” Id. at 165. The record amply supports the IJ’s findings. Petitioner’s own testimony — which is the only evidence we have on this point — makes it plain that the soldier was “not interested” in whether petitioner was still “politically involved,” id. at 161; his only concern was with petitioner’s ability to pay.

Our cases have long held that private acts of extortion cannot form the predicate for an asylum claim. In Bolshakov v. INS, 133 F.3d 1279 (9th Cir.1998), we held that applicants seeking to prove past persecution must show more than merely “that they had been the victim of criminal activity.” Id. at 1281 . Similarly, in Florez-de Solis v. INS, 796 F.2d 330 (9th Cir.1986), we held that a group that collected a private debt violently was not acting on account of political opinion despite its political affiliation. Id. at 355 ; see also Sangha, 103 F.3d at 1491 (holding that recruitment of petitioner by a guerilla group may have had non-political justifications).

The majority disregards these cases and purports to follow Gonzales-Neyra v. INS, 122 F.3d 1293 (9th Cir.1997). But this case cuts entirely the wrong way. Gonzales-Neyra was subjected to extortion by guerrillas trying to raise funds to support their cause. After paying for a time, he stopped because he objected to their political agenda, at which point the guerrillas stepped up their harassment. We held that the guerrillas’ action was politically motivated because it was taken in retaliation for petitioner’s political disagreement: Petitioner’s “life and business [were] threatened only after he expressed his political disagreement with the guerilla organization, and only after he made clear *1004 that his refusal to make further payments was on account of that disagreement.” Id. at 1294 (emphases added). We explained that “[t]he persecution of which Gonzales-Neyra complains is not the extortion, but the threats upon his life and business that were made after the guerillas learned of his political orientation.” Id. at 1296 (emphases added).

By contrast, the IJ here found that petitioner’s “reason for not paying the soldier was not motivated by his political opinion or lack thereof, but rather by his inability to pay. The soldier did not appear motivated by political interests, but rather by purely personal and economic interests.” A.R. at 165. The IJ continued: “The soldier did not appear to interpret the applicant’s inability to pay as politically motivated, or as an indication that he still was politically involved with the Mojahedeen.” Id. The IJ then found that “[t]he actions of the soldier were not directed toward modifying or punishing the applicant’s previous political opinion.” Id. at 166. Finally, the IJ concluded that “the soldier’s actions were extortion related, motivated not by [petitioner’s] political opinion, but rather motivated by his ability to pay.” Id. The Id’s findings could not have been clearer.

The majority nevertheless reverses on the curious ground that the soldier’s motive in blackmailing petitioner was “inextricably intertwined with the Petitioner’s past political affiliation.” Maj. op. at 1000. It is true, as the majority asserts, that there may be more than one motive for persecution. The IJ recognized this when she noted that “[i]n some cases, possible mixed motives for inflicting harm exist.” A.R. at 163. Nevertheless, the IJ found that the soldier here did not act from mixed motives, but only to enrich himself: The soldier was “motivated by ... purely personal and economic interests.” Id. at 165 (emphasis added). The majority holds that the IJ was required to find a mixed motive, even though she was convinced that the persecutor’s motive was not mixed. The majority thus substitutes a new rule of law for a finding of fact, precisely the kind of maneuver the Supreme Court disapproved in Ventura, 123 S.Ct. at 356 .

It is, moreover, a rule of law with sweeping implications. Political belief is only one ground for asylum; there are a number of others, such as religion and ethnicity. See n. 2 supra. If blackmailing someone on grounds of political opinion or imputed political opinion is a basis for asylum, the same would have to be true of blackmail on any of the other grounds specified by the asylum statute. Thus, if someone in a Muslim country is blackmailed for having failed to wear proper face covering or for drinking alcohol, the blackmail automatically becomes persecution “on account of’ religion for purposes of asylum, even if the blackmailer is interested only in money. This vastly and unjustifiably expands the grounds for asylum beyond those contemplated by Congress.

b. Equally unfounded is the majority’s conclusion that the blackmailer acted on behalf of his government. The record amply supports the IJ’s contrary finding that the blackmailer was acting for “his own personal monetary gain” and “not in a government capacity.” A.R. at 166. The record contains no evidence that the Iranian government was aware of the soldier’s activities, much less encouraged him. Indeed, the fulcrum of the extortion threat was that the soldier would not report petitioner to the government. Moreover, the IJ found that petitioner and his family “were permitted free departure from Iran,” and petitioner “was not required to obtain an exit permit,” even though “the Iranian government requires exit permits for citizens it considers politically suspect.” Id. at 165. On this record, the trier of fact was entitled to find that the persecutor, *1005 though on the government payroll, was on a frolic of his own.

The majority once again disregards the IJ’s perfectly reasonable finding that petitioner “does not appear to have been acting in concert with the government,” id. at 166, by displacing it with a rule of law — a rule plucked from a wholly different context, no less. Maj. op. at 1000 (citing Screws v. United States, 325 U.S. 91 , 65 S.Ct. 1031 , 89 L.Ed. 1495 (1945)). Putting aside whether it makes any sense to apply caselaw dealing with the conduct of American police officers to an Iranian soldier’s relationship to his government, the majority patently misapplies Screws . The very point of Screws was that the defendants there were acting as law enforcement officers carrying out official duties. Screws, 325 U.S. at 111 , 65 S.Ct. 1031 . Because “[i]t was their duty under Georgia law to make the arrest effective,” their conduct, though unauthorized by state law, still “comes within the statute.” Id. at 107-08 , 65 S.Ct. 1031 . But the Court limited its analysis to the “official” conduct of government agents — conduct calculated to achieve law enforcement purposes. Id. at 111 , 65 S.Ct. 1031 . The Court made a point of saying that “acts of officers in the ambit of their personal pursuits are plainly excluded.” Id.

Here we have a finding that the blackmailer was acting out of his own private interests, not as a representative of his government, nor to advance some governmental purpose. The record amply supports that finding; indeed, there is no contrary evidence. The majority’s conclusion thus cannot be based on the particular facts and circumstances of this case, because none support the majority’s view. Instead, the majority adopts a new rule of law that the actions of agents and employees of foreign governments are always attributable to that government, even if the employees are acting entirely without official sanction and are pursuing only personal objectives.

This is a rule with explosive potential. Governments all over the world have a multitude of agents and employees, especially if one includes every rank-and-file member of the armed forces — like the soldier who blackmailed petitioner. Moreover, corruption by government officials— use of their official position to line their pockets — is a way of life in many countries. According to today’s decision, all such conduct is automatically attributable to the government employing the corrupt official — at least if the issue is presented in the Ninth Circuit. I doubt this is what Congress had in mind when it gave the courts of appeals authority to review asylum petitions.

2. The majority also seems to hold that petitioner is eligible for asylum based on the alternative prong of the test, namely that petitioner established a well-founded fear of future persecution, whether or not he showed past persecution. 4 Once again, the IJ made specific findings the majority entirely ignores. The IJ found that “no accounts of extortion, threats or persecution against former Mojahedeen members have been reported.” A.R. at 167. Though petitioner claimed that stories of government persecution had been reported in Iranian newspapers published in the United States, the IJ did not credit this argument because petitioner “failed to provide *1006 the Court with even one of these newspaper articles.” Id. Furthermore, the IJ found that, although the soldier knew where petitioner’s parents and sisters lived, they had “continued to reside in Iran without any contact from, or harassment by, this soldier or any government official for nearly the past 9 years.” Id. Based on these facts, the IJ found that petitioner had no legitimate reason to fear government persecution upon his return to Iran.

The majority disregards these findings and, instead, takes refuge in a long and irrelevant passage from the 1997 Iran Country Report. Maj. op. at 1000. The passage shows that conditions in Iran, especially in prison, were pretty bad at the time, but it says nothing of consequence concerning petitioner’s situation. Nor is the majority’s position supported by the Country Report passage it quotes earlier in its opinion, discussing actions taken against certain members of the Mojahe-deen. Id. at 994. The report explains that the Iranian government has targeted leaders and prominent members of the Mojahedeen because of that group’s terrorist activities, which “includ[ed] assassinations and car bombings” and, during the Iran-Iraq War, “full-scale military operations against the Islamic regime.” A.R. at 125. That Iran targeted leaders of an organization actively involved in terrorism is hardly surprising; our government does much the same. This does not mean that Iran is persecuting all those ever affiliated with the group, no matter how long ago or how peripheral their involvement. Petitioner was not even a member of the Moja-hedeen and his activities were limited to distributing literature. See page 1002 & n. 1 supra. There is nothing in the record to support the finding that people in petitioner’s position were being persecuted.

Grasping at straws, the majority refers to petitioner’s friend Cyrus Yaghley, who supposedly was blackmailed and eventually disappeared. Maj. op. at 999; see also id. at 998. But Yaghley’s name never appears in petitioner’s testimony; here is all petitioner says that is even remotely relevant:

A.R. at 210-11 (emphases added). Petitioner says nothing about who these individuals were, whether they were involved with the Mojahedeen and how they came to be in jail, though he does admit that he only “sort of’ knew “of’ them — hardly how one would refer to a friend. It’s conceivable that this testimony refers to Yaghley but, if so, its very vagueness makes it useless to petitioner, and the IJ was fully justified in discounting its significance. 5

*1007 The simple fact is that the majority provides no justification at all for reversing the IJ’s and the BIA’s finding that petitioner lacks a well-founded fear that he would be persecuted were he to return to Iran. The majority’s perfunctory analysis not only does not compel reversal, it doesn’t even support it. See INS v. Elias-Zacarias, 502 U.S. 478 , 481 n. 1, 112 S.Ct. 812 , 117 L.Ed.2d 38 (1992). All it does is to further “whittle away the authority and discretion of immigration judges and the BIA.” Abovian v. INS, 257 F.3d 971, 971 (9th Cir.2001) (Kozinski, J., dissenting from denial of rehearing en banc).

3. Which brings us back to who is the master in asylum cases. The Supreme Court’s summary reversal in two immigration cases last Term must be understood as a wake-up call that our jurisprudence in this area of the law is in need of repair. See Ventura, 123 S.Ct. at 355-56 ; Chen, 123 S.Ct. at 549 . The Supreme Court does not normally take up routine cases involving simple errors in settled areas of the law. But the Supreme Court did so in these cases upon the urging of the Solicitor General, who pointed out the “increasing[] importan[ee]” of asylum cases “to enforcement of the immigration laws.” Chen, Pet. for Cert, at *28. Moreover, as the Solicitor General noted, “[ajsylum decisions ... are ‘vitally and intricately interwoven with contemporaneous policies in regard to the conduct of foreign relations, the war power,’ and the definition of the national community.” Id. (quoting Harisiades v. Shaughnessy, 342 U.S. 580, 588-89 , 72 S.Ct. 512 , 96 L.Ed. 586 (1952)).

The Supreme Court pointedly quoted the Solicitor General’s complaint that, in an area where a uniform national policy is crucial, the Ninth Circuit is often “in conflict with other courts of appeals, which generally respect the BIA’s role as fact-finder.” Ventura, 123 S.Ct. at 355 (citing INS v. Ventura, Pet. for Cert. at 11, available at http://www.usdoj.gov/osg/briefs/ 2002/2pet/7pet/2002-0029.pet.aa.pdf, and Chen, Pet. for Cert. at **22-24) (internal quotation marks omitted). The Solicitor General’s petitions in Ventura and Chen actually said quite a bit more. For example, we were accused of “developing] a body of circuit law that relieves the applicant of his burden of proof in asylum cases and allows the court to substitute its own views about contested record evidence for reasonable determinations of the BIA.” Chen, Pet. for Cert, at *14. Our precedents were also characterized as “absurd! ],” id. at *18, and as “defiying] the most basic rules of judicial review,” id. at *13. The Supreme Court tactfully spared us the embarrassment of quoting these passages, but how much longer can we count on such forbearance?

Having identified our proclivity for error, the Court proceeded to administer a mini-tutorial as to the applicable black-letter principles of administrative law. Ventura, 123 S.Ct. at 355-56 . While the Court was careful to limit its ruling to the facts presented, its message to us was clear to anyone with eyes to see: Stop substituting your judgment for that of the BIA; give proper deference to administrative factfinding; and do not adopt rules of law that take away the agency’s ability to do its job. In other words, stop fiddling with the agency’s decisions just because you don’t like the result.

We could, of course, read the Supreme Court’s decisions in Ventura and Chen as limited to the questions presented in those *1008 cases. But this would be a big mistake. The Court gave us fair warning that our jurisprudence in this area of the law falls well outside the mainstream. The Court also gave us a gentle hint that we must revise our mindset on the key question about who’s the master when it comes to immigration cases. We must come to understand and accept — as the other courts of appeals have — that “[wjithin broad limits the law entrusts the agency to make the basic asylum eligibility decision.” Id. at 355. The majority’s reversal of the BIA’s and IJ’s perfectly reasonable, thoroughly considered decisions in this case shows that my colleagues have not yet taken this lesson to heart.

. This is entirely consistent with petitioner's testimony:

. The statute provides other grounds that might be the basis of persecution, such as religion or ethnicity, but petitioner has not claimed relief based on any of these. 8 U.S.C. § 1101 (a)(42)(A).

. Petitioner may also meet the statutory criteria by showing persecution by a non-governmental entity the government is “unable or unwilling” to control. See Singh v. INS, 134 F.3d 962 , 967 n. 9 (9th Cir.1998). Petitioner has not claimed eligibility on the ground that the government was “unable or unwilling” to control the soldier, nor can he, as there is no indication the Iranian government was even aware of the soldier's activities.

. Admittedly, the majority opinion is somewhat confused on this point. The footnote pertaining to past persecution is, for reasons that are unclear, placed within the paragraph discussing future persecution. See Maj. op. at 1000 n. 3. But, the majority holds that petitioner has established eligibility for asylum "on both grounds specified in the statute,” id. at 998 , and there would be no point in discussing future persecution if the majority were only relying on the presumption arising from past persecution.

. Yaghley’s name does appear in petitioner's asylum application. A.R. at 249. But the IJ’s credibility finding covers only petitioner's live testimony, not assertions in his application. In any event, the application hardly helps petitioner. He states there as follows: "A similar incident had occured [sic] to a friend, Cyrus Yaghley of Tehran; he was similarly blackmailed by the Pastars, didn't pay, and was imprisoned with his entire family in Tehran in 1987 and has not been heard of since.” Id. Petitioner says nothing about whether Yaghley was ever involved with the Mojahe-deen and, if so, whether his level of involvement was similar to petitioner's. Nor does petitioner explain how he knows the circumstances surrounding Yaghley's disappearance. If petitioner "sort of” knew Yaghley or knew only "of” him, id. at 210-11, his information on these important points would be based entirely on hearsay and rumor. Petitioner could have filled out these details when he testified, but he did not; his testimony is vaguer even than his application. Petitioner *1007 bears the burden of establishing his entitlement to asylum; his failure to provide any details supporting his claim that individuals involved in the distant past with the Mojahe-deen have been targeted for persecution provides ample basis for the IJ’s decision to reject his testimony on this point. Id. at 167.