I. BACKGROUND

Nikolay Angov, a Bulgarian citizen, claims he was persecuted by the Bulgarian government because he is Roma.1 He alleges repeated abuse at the hands of the Bulgarian police, including beatings, false accusations of crimes and illegitimate arrests. After three years of this treatment, he fled Bulgaria and sought asylum in the United States.

An IJ conducted asylum hearings in early 2004, during which Angov presented several documents, including two Bulgarian subpoenas that ordered him to appear at a Sofia police station. The immigration judge (“IJ”) allowed the government to obtain a State Department investigation of Angov’s allegations. See 8 C.F.R. § 208.11. The investigation was conducted by our consulate in Sofia, and

1 Angov’s brief refers to him as “Roma” or “gypsy” interchangeably. So do we.

ANGOV V. HOLDER 5

the results were summarized in a letter signed by Cynthia Bunton, Director of Department of State’s Office of Country Reports and Asylum Affairs.

The IJ admitted the Bunton Letter, which stated that the Embassy had contacted “an official in the Archive Department at the 5th Police District in Sofia.” The official found a number of errors in the subpoenas, suggesting that they were forgeries: (1) Three officers named in the subpoena—Captain Donkov, Lieutenant Slavkov and Investigator Vutov—never worked for the police department; (2) the case and telephone numbers were wrong; and (3) although the subpoenas mentioned room 4 on the second floor of the department and room 5 on the first floor, there are no rooms by those numbers. The official also explained (4) that the seal on the subpoena was too small.

Bunton also stated that the embassy investigator (5) was unable to locate Angov’s claimed past residences; and (6) that the neighborhood where Angov lived was only twenty to thirty percent Roma, though Angov claimed that he lived in a “gypsy neighborhood.” Attached to the letter were five photographs of the places the investigator had visited while trying to verify the addresses.

Angov’s industrious lawyer submitted a plethora of rebuttal evidence, including photos, maps, an article about Angov’s neighborhood and a letter apparently signed by someone named Daniela Mihaylova, who identified herself as the legal programs director of a Roma human rights organization in Bulgaria. Angov also argued that, without the opportunity to cross-examine the investigator, the admission of the Bunton Letter would violate his statutory and constitutional rights. In response to Angov’s objection, the government attorney asked the State Department to produce an employee to testify about the investigation. State responded with a letter authored by Nadia Tongour, Bunton’s successor. The Tongour Letter provided some general background information on State’s investigation procedures, but explained that it’s State’s policy to refrain from providing further specific information about an overseas investigation.

Based on the Bunton Letter, the IJ made an adverse credibility finding and denied Angov’s applications for asylum, withholding of removal and relief under the Convention Against Torture. The Board of Immigration Appeals (“BIA”) adopted and affirmed the IJ’s ruling denying relief, and his determination that the subpoenas are fraudulent. The BIA also denied Angov’s motion to supplement the record with a recent Sixth Circuit opinion that Angov claimed constituted new evidence of a “pattern and practice” of law-breaking by officials in the Sofia consulate. See Alexandrov v. Gonzales, 442 F.3d 395 (6th Cir. 2006).

II. ANALYSIS

A. Motion to Remand

Angov claims the BIA abused its discretion by denying his motion. See Movsisian v. Ashcroft, 395 F.3d 1095, 1098 (9th Cir. 2005). His brief before the BIA spent just two sentences explaining this argument:

Respondent respectfully submits a copy of Alexandrov v. Gonzales to supplement the record in this case. The document is submitted to document a pattern and practice

ANGOV V. HOLDER 7

of procedural and substantive violations of the law and applicable regulations by the consulate in Sofia during overseas investigations and in divulging the identity of asylum applicants to the authorities in Bulgaria in violation of C.F.R. 208.6 [sic].

“Since a motion to remand is so similar to a motion to reopen, the motion to remand should be drafted in conformity with the regulations pertinent to motions to reopen . . . .” Rodriguez v. INS, 841 F.2d 865, 867 (9th Cir. 1988) (internal quotation marks omitted). The applicable regulation provides that a motion to reopen shall state “the new facts that will be proven at a hearing to be held if the motion is granted” and be supported by affidavits or other “evidentiary material.” 8 C.F.R. § 1003.2(c)(1). But Angov didn’t provide any evidence supporting his motion nor did he even explain why he believed that section 208.6 had been violated.2 The BIA did not abuse its discretion in denying Angov’s motion to remand.

B. Admission of the Bunton Letter

Angov claims that the admission of, and the IJ’s and BIA’s reliance on, the Bunton Letter violated his statutory and constitutional rights. See 8 U.S.C. § 1229a(b)(4)(B);

2 8 C.F.R. § 208.6(a) provides that “[i]nformation contained in or pertaining to any asylum application . . . shall not be disclosed without the written consent of the applicant.” Angov argues that Alexandrov “exposed the improprieties that have riddled overseas investigations in the Sofia consulate,” including that investigations were often conducted by foreign service nationals, that someone other than a consular officer could have authored embassy reports and that consular officials often signed reports written by others. None of these arguments were presented to the BIA. 8 C.F.R. § 1240.10(a)(4); Cinapian v. Holder, 567 F.3d 1067, 1074–75 (9th Cir. 2009). In considering Angov’s argument, we review the IJ’s decision, except for the portion that the BIA didn’t clearly adopt—here, the IJ’s conclusion that the Department of State’s inability to verify Angov’s addresses supported an adverse credibility finding. See Joseph v. Holder, 600 F.3d 1235, 1239–40 (9th Cir. 2010). On that issue, we review the BIA’s decision.

While we review legal questions de novo, “[t]he BIA’s interpretation and application of the immigration laws are generally entitled to deference.” Hernandez-Mancilla v. Holder, 633 F.3d 1182, 1184 (9th Cir. 2011); Zetino v. Holder, 622 F.3d 1007, 1011–12 (9th Cir. 2010). The agency’s factual findings—such as its adverse credibility determination—are reviewed for substantial evidence and can be reversed only if the evidence “compels” a contrary conclusion. See Rizk v. Holder, 629 F.3d 1083, 1087–88 (9th Cir. 2011) (emphasis omitted).

(i) Due Process

Angov claims that the IJ’s reliance on the Bunton Letter violated his constitutional right to procedural due process. But Angov has no such right. He is an alien who has never formally entered the United States. He presented himself at the San Ysidro port of entry without valid entry documents and sought asylum. “[A]n alien seeking admission has not ‘entered’ the United States, even if [he] is in fact physically present.” Kwai Fun Wong v. United States, 373 F.3d 952, 971 (9th Cir. 2004). “[O]ur immigration laws have long made a distinction between those aliens who have come to our shores seeking admission . . . and those who are within the United States after an entry.” Leng May Ma v. Barber,

ANGOV V. HOLDER 9

357 U.S. 185, 187 (1958). Aliens “who have once passed through our gates, even illegally,” are afforded the full panoply of procedural due process protections, and “may be expelled only after proceedings conforming to traditional standards of fairness.” Shaughnessy v. United States ex rel. Mezei, 345 U.S. 206, 212 (1953). But those, like Angov, who have never technically “entered” the United States have no such rights. Id. For Angov, procedural due process is simply “[w]hatever the procedure authorized by Congress” happens to be. Id. (internal quotation marks omitted); see also Landon v. Plasencia, 459 U.S. 21, 32 (1982) (“[A]n alien seeking initial admission to the United States requests a privilege and has no constitutional rights regarding his application . . . .”).

Angov’s claim of a procedural due process violation simply can’t be squared with the Supreme Court’s teachings in Mezei and Landon, nor with our circuit’s settled precedent. See Barrera-Echavarria v. Rison, 44 F.3d 1441, 1449 (9th Cir. 1995) (“[E]xcludable aliens have no procedural due process rights in the admission process . . . .”).3

3 We note that four circuits have held that reliance on documents like the Bunton Letter in asylum proceedings violates due process. See Banat v. Holder, 557 F.3d 886, 892–93 (8th Cir. 2009); Anim v. Mukasey, 535 F.3d 243, 256–58 (4th Cir. 2008); Alexandrov, 442 F.3d at 407; Ezeagwuna v. Ashcroft, 325 F.3d 396, 405–08 (3d Cir. 2003). Because Angov does not have a constitutional right to procedural due process, that question is not before us. We also note that two other circuits have held that asylum applicants like Angov are entitled to certain “minimum due process” rights in the application of their statutory rights. See Marincas v. Lewis, 92 F.3d 195, 203–04 (3d Cir. 1996); Augustin v. Sava, 735 F.2d 32, 37 (2d Cir. 1984); see also Meachum v. Fano, 427 U.S. 215, 226 (1976). Whether asylum applicants are owed such “minimum due process” is an open question in our circuit, but it is not one we need to resolve here. Angov was clearly given fair access to all his statutory rights. What he asks for instead are due process protections that go beyond those which Congress (ii) Statutory Rights

Angov’s challenge to the admission of the Bunton Letter is therefore purely statutory. In assessing such a challenge, we must first ask whether the IJ made legal error by denying Angov any of his statutory rights. Angov claims that he was denied his right to examine evidence against him. See 8 U.S.C. § 1229a(b)(4)(B). But the record tells a different story. He was allowed to examine the Bunton Letter, and given ample time to produce substantial evidence to rebut it. See p. 5 supra; cf. Cinapian, 567 F.3d at 1076 (had the government given petitioners a chance to examine forensic reports before hearing, they may have been able to produce rebuttal evidence).

Angov also argues that he was denied his statutory right to cross-examine the witnesses against him. We’ve held that, before hearsay statements made by an absent witness can be admitted into an immigration hearing, “‘the government must make a reasonable effort . . . to afford the alien a reasonable opportunity to confront the witnesses against him or her.’” Hernandez-Guadarrama v. Ashcroft, 394 F.3d 674, 681 (9th Cir. 2005) (quoting Saidane v. INS, 129 F.3d 1063, 1065 (9th Cir. 1997)); see also § 1229a(b)(4)(B); Baliza v. INS, 709 F.2d 1231, 1234 (9th Cir. 1983).

The government is, of course, not required to produce Bulgarian police officials at an immigration hearing in the United States. Such a requirement would make it virtually impossible for the government to introduce evidence rebutting an alien’s claims relating to conduct abroad.

has provided him. But, as an alien who has never entered the United States, those protections are unavailable to him.

ANGOV V. HOLDER 11

Instead Angov, and the dissent, claim that the immigration authorities should have obtained a witness from the Department of State to verify the letter’s contents. But hauling State Department officials into court wouldn’t ameliorate the dissent’s concerns, because the letters they author inescapably rely on foreign officials who aren’t amenable to cross-examination.

In any event, the government here did make a reasonable effort to obtain a State Department witness, but was prevented from doing so by State’s policy of not releasing follow-up information regarding its overseas investigations. The dissent claims that “allowing one executive branch agency to rely on another executive branch agency’s blanket policy of refusing to provide certain information is tantamount to granting the government the kind of unfettered discretion we repudiated in Baliza.” But Baliza offers no support to the dissent’s position. There the government relied on the affidavit of an alien’s ex-wife as the basis for a fraudulent marriage charge while barely even trying to investigate her whereabouts. The declarant was not a government official and the government gave no justification for failing to find her. By contrast, the declarant here is a government official speaking in her official capacity. And the immigration authorities’ decision not to present her in person was made pursuant to a coordinate department’s reasonable policy governing the secrecy and safety of its officers. Because neither the immigration authorities nor the State Department acted unreasonably in failing to compel Bunton to testify, Angov’s statutory rights were not violated. (iii) Substantial Evidence

Because the IJ did not erroneously deny Angov a statutory right, our review is limited to whether the IJ’s adverse credibility finding was supported by substantial evidence. “This strict standard bars a reviewing court from independently weighing the evidence,” and requires us to “deny the Petition unless Petitioner [has] presented evidence so compelling that no reasonable factfinder could find that Petitioner” was not credible. Singh v. INS, 134 F.3d 962, 966 (9th Cir. 1998) (internal quotation marks omitted).

Despite the generally flexible—and highly deferential— nature of substantial evidence review, Angov appears to argue for a per se rule under which immigration judges must blind themselves to the findings of a State Department letter, unless it provides particular information regarding how an investigation was conducted. Surprisingly, Angov’s radical proposal accords with the view of the Second Circuit, which has held that a document akin to the Bunton Letter was “inherently unreliable” because it didn’t reveal the qualifications of the investigator, the extent of the investigation or the methods used to verify the information. Lin v. U.S. Dep’t of Justice, 459 F.3d 255, 271–72 (2d Cir. 2006). Under Second Circuit law, therefore, documents like the Bunton Letter categorically “cannot support [an] adverse credibility finding.” Id. at 272. We reject this approach. Substantial evidence review requires an appellate court to consider the reasonableness of an agency’s conclusions; it does not empower us to craft quasi-statutory criteria governing the admissibility of evidence in agency proceedings. In light of our departure from the holding of a sister circuit—one with the second-largest immigration

ANGOV V. HOLDER 13

docket in the country—we offer a thorough explanation for our rationale.

1. Congress and the Attorney General have accorded aliens like Angov a variety of procedural rights, including the right to be present at the hearing; to be represented by counsel; to examine the evidence against him and present counter-evidence; to cross-examine witnesses; and to have a written record kept of the proceedings. 8 U.S.C. § 1229a(b)(4). But neither the statute nor the regulations give the asylum applicant a right to a particular quality of the evidence presented against him. Instead, he is given the right to have an impartial adjudicator assess the evidence. When exercising grace towards individuals entitled no procedural rights under the constitution, Congress can set the precise limits of what it grants and what it withholds. That then defines the process an asylum seeker like Angov is due.

With that in mind, let’s put Angov’s claims into some context. The IJ found that Angov presented forged documents. This is a serious matter that, if true, should not merely result in the immediate termination of Angov’s asylum petition, but also in criminal prosecution for immigration fraud. But the IJ wasn’t fazed by discovery of the fraud; he went on to decide whether Angov’s asylum claim could be sustained despite the forgeries. No other adjudicator in the United States would react with such equanimity to finding that a party had tried to bamboozle it.

This points to an unfortunate reality that makes immigration cases so different from all other American adjudications: Fraud, forgery and fabrication are so common—and so difficult to prove—that they are routinely tolerated. Our circuit is no exception. See Abovian v. INS, 257 F.3d 971 (9th Cir. 2001) (Kozinski, J., dissental).

The reason for this deplorable state of affairs is not difficult to figure out. The schizophrenic way we administer our immigration laws creates an environment where lying and forgery are difficult to disprove, richly rewarded if successful and rarely punished if unsuccessful. This toxic combination creates a moral hazard to which many asylum applicants fall prey.

First, the reward: the opportunity to be lawfully admitted into the United States. Those born with U.S. citizenship cannot imagine what this is worth to the world’s poor and oppressed billions, most of whom would come here tomorrow if they could. Gaining a lawful foothold in America is an incalculable benefit. It sets an immigrant on the path to a peaceful life in a free society, economic prosperity, citizenship and the opportunity to bring family members in due course. A prize like this is worth a great deal of expense and risk. Telling an elaborate lie, and coming up with forged documents and mendacious witnesses to back it up, is nothing at all when the stakes are so high.

And the risk of getting caught is low. As eight members of this court pointed out in Abovian:

The specific facts supporting a petitioner’s asylum claim—when, where, why and by whom he was allegedly persecuted—are peculiarly within the petitioner’s grasp. By definition, they will have happened at some time in the past—often many years ago—in a foreign country. In order for the INS to

ANGOV V. HOLDER 15

present evidence “refuting or in any way contradicting” petitioner’s testimony, it would have to conduct a costly and often fruitless investigation abroad, trying to prove a negative—that the incidents petitioner alleges did not happen.

257 F.3d at 976. There’s very little the United States can do to investigate obscure incidents that allegedly occurred in countries on the other side of the globe. Even if it were economically feasible, we can’t send the FBI into a foreign country to conduct a full field investigation. The best we can do is to have consular personnel check basic facts, in addition to the many other functions they perform. And we have very few U.S. consular personnel on the ground in most countries; in all of Bulgaria, there are fewer than two dozen. See U.S. Sec’y of State, 1 Congressional Budget Justification, Department of State Operations, Fiscal Year 2013, at 306 (2012). All told, there are fewer than 6000 consular officials in embassies and consulates spread out across more than 170 countries. Id. at 227–311.

Finally, if an alien does get caught lying or committing fraud, nothing very bad happens to him. Sure, he may be ordered removed, but most aliens who aren’t in custody remain here long after their removal orders become final. See, e.g., Office of the Inspector Gen., U.S. Dep’t of Justice, The Immigration and Naturalization Service’s Removal of Aliens Issued Final Orders iii (2003) (reporting that “the INS removed only 3 percent of nondetained asylum seekers with final removal orders”); see also Mark Hamblett, Circuit Sets Policy for Removal Cases Deemed Low Priority by U.S., N.Y. L.J., Oct. 18, 2012 (discussing policy that calls for “the exercise of prosecutorial discretion to focus removal efforts on the most high-priority cases”). And if they do get sent back—at our expense—what’s lost? They wind up where they started. Would-be immigrants almost never get prosecuted for presenting forged documents in support of asylum petitions, unless they commit some additional misconduct. See, e.g., United States v. Jawara, 474 F.3d 565, 570 (9th Cir. 2007) (defendant charged with document fraud and conspiracy to commit marriage fraud). Consequently, immigration fraud is rampant.

Take, for instance, Angov’s compatriot, Pavel Pavlov. Pavlov sought asylum as a persecuted gypsy, just like Angov. They even have the same lawyer. But Pavlov’s story took a different turn when his wife gained U.S. citizenship and he sought adjustment of status. In the process, he had to disclose that his asylum application was a tissue of lies. Specifically, Pavlov admitted that he wasn’t persecuted in Bulgaria. In fact, he’s not even a gypsy.

Americans galore wind up in federal prison every year for far less significant lies on government forms or bank loan applications. See, e.g., United States v. Prince, 647 F.3d 1257, 1260–61, 1265 (10th Cir. 2011); United States v. Sandlin, 589 F.3d 749, 751–53 (5th Cir. 2009); United States v. Jack, 216 F. App’x 840, 841–43 (11th Cir. 2007). So was Pavlov appealing his criminal conviction? Certainly not. The BIA barred Pavlov from obtaining any relief under our immigration laws because he had filed a frivolous (read: fraudulent) asylum petition—a decision he had the chutzpah to appeal. See Pavlov v. Holder, 697 F.3d 616 (7th Cir. 2012).

Cases involving fraudulent asylum claims are distressingly common. See, e.g., Cheema v. Holder, 693 F.3d

ANGOV V. HOLDER 17

1045, 1046–47 (9th Cir. 2012); Dol v. Holder, 492 F. App’x 774, 775 (9th Cir. 2012); Zheng v. Holder, 672 F.3d 178, 180–81 (2d Cir. 2012); Fernandes v. Holder, 619 F.3d 1069, 1074–76 (9th Cir. 2010); Ghazali v. Holder, 585 F.3d 289, 290–91 (6th Cir. 2009); Ribas v. Mukasey, 545 F.3d 922, 925–26 (10th Cir. 2008); Siddique v. Mukasey, 547 F.3d 814, 815–16 (7th Cir. 2008); Rafiyev v. Mukasey, 536 F.3d 853, 855–57 (8th Cir. 2008); Dhital v. Mukasey, 532 F.3d 1044, 1047–48 (9th Cir. 2008) (per curiam); Chen v. Mukasey, 527 F.3d 935, 938–39 (9th Cir. 2008); Ahir v. Mukasey, 527 F.3d 912, 914–16 (9th Cir. 2008). And for every case where the fraud is discovered or admitted, there are doubtless scores of others where the petitioner gets away with it because our government didn’t have the resources to expose the lie.

The Second Circuit has given this already shaky system a swift kick in the gut. As we explain further below, its ruling makes it pretty much impossible for the immigration authorities to carry out even the little bit of fact checking they now manage to do. Its decision smothers the State Department’s informal process of checking up on asylum petitions in layers of procedural complexity that will prove impossible to administer in practice. Perhaps the Supreme Court or Congress will intervene and decide who’s right.

2. The basic question that confronts us is this: In a system where there are pervasive, structural incentives for fraud, are we to disable our triers of fact from considering certain evidence—which may be essential to weeding out fraudulent claims—when that evidence lacks particular details that may bear on its reliability? Remember, the Second Circuit regards documents like the Bunton Letter as inherently unreliable. We need not—and do not—conclude that such letters will always lead to adverse credibility findings; we simply disclaim the conclusion that they must be excluded from an immigration judge’s consideration when they fail to provide sufficient identifying detail.

We acknowledge the Bunton Letter lacks certain indicia of reliability, but we cannot say, under our “extremely deferential” review, that its use alone constitutes grounds to reverse the IJ’s adverse credibility determination. Wang v. INS., 352 F.3d 1250, 1257 (9th Cir. 2003). First of all, Angov has the burden of proving his eligibility for asylum. See 8 C.F.R. § 1208.13(a). The government has no burden; it can present evidence solely to rebut or impeach petitioner’s case. The IJ and the BIA could reasonably conclude that the Bunton Letter is at least sufficient to cast doubt on Angov’s evidence and force him to come up with more solid proof to support his claim.

Angov finds fault with the Bunton Letter because it “provides no information as to who conducted the investigation; who obtained, stored and verified the information underlying the conclusion expressed in the document [or] when and under what authority the investigation was conducted.” He notes that the Bunton Letter offers no explanation for many of its conclusions—for example, that both the case numbers and the telephone numbers listed on the fraudulent subpoenas were incorrect. These are all interesting points to raise at the hearing, and the absence of a satisfactory response from the government might well convince the trier of fact to disregard the letter. But in this instance, the IJ, in his discretion, chose to credit the letter. That was his prerogative, and our review is limited to whether that decision is permissible in light of the evidence.

ANGOV V. HOLDER 19

The doubts as to the letter’s reliability flow from the fact that the rules of evidence, and the hearsay rules in particular, don’t apply to administrative proceedings. See Richardson v. Perales, 402 U.S. 389, 400–02 (1971); Hernandez- Guadarrama v. Ashcroft, 394 F.3d 674, 681 (9th Cir. 2005). This inevitably leaves some uncertainty that would be eliminated if this were a formal trial subject to the rules of evidence. But it doesn’t deprive the opposing party of any and all means of rebutting the hearsay declarant’s assertions.

The Bunton Letter does come to certain factual conclusions: that the addresses identified by Angov in his asylum petition don’t exist; that the officers—Captain Donkov, Lieutenant Slavkov and Investigator Vutov—and room numbers specified in the subpoenas presented by Angov don’t exist; that the seals on the subpoenas are the wrong size; and that the part of the city where Angov claimed to live was only twenty to thirty percent Roma. Each of these assertions describes facts in the real world, so it’s possible to rebut Bunton by presenting proof that those facts are not as the Bunton Letter describes them.

In fact, Angov did precisely that with respect to the two addresses. He presented a letter from someone in Bulgaria, who explained that the Bunton Letter’s conclusions about the addresses are wrong. See p. 5 supra; Appendix. And the BIA seems to have been swayed, as it noted that the “record is unclear” about whether Angov was telling the truth about the addresses.

Angov was free to present similar evidence to undermine the Bunton Letter’s statements about the subpoenas. He could have had Ms. Mihaylova from the human rights organization or some other friend in Sofia visit the police station and try to find out whether the rooms referenced in the Bunton Letter do or don’t exist. He might also have been able to obtain a roster of the names of police officials in Sofia and shown that it contains the names of the officers referenced in the subpoenas.

The Bunton Letter also asserts that the phone numbers in the subpoenas aren’t correct. Angov or one of his friends could have called the numbers and asked whether he’d reached the police station—and then submitted an affidavit to that effect. The same is true about the seals: Angov or his friends might have tried to obtain an official copy of the police seal from the Bulgarian government and introduced it into evidence. He did none of these things, perhaps because he knew that the subpoenas were forged.

Where the petitioner has the burden of proof, there’s nothing unfair about having a U.S. government agent check out some of his basic facts and inform the IJ of possible discrepancies. This forces the petitioner to obtain further evidence supporting the challenged claims. There might be situations where obtaining further evidence is impossible, such as where the petitioner has fled from a closed society and can find no one willing or able to obtain the evidence he needs. In such cases, we don’t hold the petitioner’s failure to present evidence against him. See Singh v. Holder, 638 F.3d 1264, 1270–71 (9th Cir. 2011). But Angov has never claimed that he couldn’t get more evidence; indeed he has resources in Bulgaria with which to do so. Based on the almost complete absence of rebuttal evidence on Angov’s part, the IJ was not unreasonable to credit the allegations in the Bunton Letter.

ANGOV V. HOLDER 21

3. There’s nothing particularly exotic about assessing an asylum applicant’s credibility by comparison with an extrinsic source. For example, the Bunton Letter’s estimate that Angov comes from a community that is only twenty to thirty percent Roma is similar to the kind of demographic estimates made by the State Department in its country reports, on which we and the BIA rely all the time. See, e.g., Dhillon v. Holder, 485 Fed. App’x 252, 253 (9th Cir. 2012); Patel v. Holder, 474 F. App’x 584, 585 (9th Cir. 2012); Sesay v. Holder, 469 F. App’x 617, 617 (9th Cir. 2012); see also Sowe v. Mukasey, 538 F.3d 1281, 1285 (9th Cir. 2008) (“U.S. Department of State country reports are the most appropriate and perhaps the best resource for information on political situations in foreign nations.” (internal quotation marks omitted)); cf. 8 U.S.C. § 1158(b)(1)(B)(iii).

Were we to hold that we can’t rely on this estimate in the Bunton Letter, we’d be casting doubt on a multitude of country reports that have no better support for their demographic estimates than the Bunton Letter. The country reports are, after all, prepared by the very same consular officials, using some of the same methods, as the Bunton Letter. See Bureau of Democracy, Human Rights & Labor, U.S. Dep’t of State, Country Reports on Human Rights Practices for 2012: Appendix A: Notes on Preparation of Reports, at 1 (2012). Indeed, Cynthia Bunton’s title when she wrote her letter was director of the Department of State’s “Office of Country Reports and Asylum Affairs.” (emphasis added). Nadia Tongour is her successor. Adopting Angov’s objection to the findings in the Bunton Letter could render country reports inadmissible in immigration proceedings.

Angov complains that the Bunton Letter might have relied on reports from foreign service nationals (FSNs). See Ezeagwuna, 325 F.3d at 406. What if it did? Our embassy in Sofia, as elsewhere, employs roughly the same number of FSNs and Americans. U.S. Sec’y of State, 1 Congressional Budget Justification, Department of State Operations, Fiscal Year 2013, at 306 (2012). Our short-staffed consular offices no doubt use FSNs, who are fluent in the local language and familiar with local conditions, to do some of the legwork. We see nothing wrong with that. Whether the investigation was conducted by U.S. citizens, FSNs or Hercule Poirot, it resulted in certain factual conclusions that can be refuted.

Submissions such as the Bunton Letter and the various country reports on which we routinely rely aren’t just a collection of statements by disconnected individuals. Rather, they are the unified work product of a U.S. government agency carrying out governmental responsibilities. As such, the report itself, and the acts of the various individuals who helped prepare it, are clothed with a presumption of regularity. See Nat’l Archives & Records Admin. v. Favish, 541 U.S. 157, 174 (2004); see also Kohli v. Gonzales, 473 F.3d 1061, 1068 (9th Cir. 2007). “[I]n the absence of clear evidence to the contrary, courts presume that [these individuals] have properly discharged their official duties.” Favish, 541 U.S. at 174 (quoting United States v. Armstrong, 517 U.S. 456, 464 (1996)).

The presumption of regularity has been applied far and wide to many functions performed by government officials. See, e.g., U.S. Postal Serv. v. Gregory, 534 U.S. 1, 10 (2001) (Post Office disciplinary procedures); United States v. Armstrong, 517 U.S. 456, 464 (1996) (prosecutorial decision making); FCC v. Schreiber, 381 U.S. 279, 296 (1965) (FCC’s decision making process); cf. INS v. Miranda, 459 U.S. 14, 16–18 (1982) (per curiam) (processing of visa application).

ANGOV V. HOLDER 23

The Bunton Letter is entitled to the presumption that those who participated in its preparation, be they FSNs, consular officers or officials at the State Department in Washington, did their jobs fairly, conscientiously and thoroughly; that each officer in the chain relied on the work of someone down the chain in whom he had confidence; that no one had a personal stake in the substance of the report; and that no one lied or fabricated evidence. Without this presumption, country reports would be no more useful than the Farmers’ Almanac or Perezhilton.com.

The dissent argues that the presumption of regularity doesn’t apply here because “[t]he key hearsay statement in the Bunton Letter comes from a Bulgarian police employee, not a U.S. government official.” But that would remain true, even with the procedural protections the Second Circuit advocates. Those protections don’t prevent Bulgarian police officers from lying, they simply make it easier for an IJ to assess the quality of investigation conducted by our consular officials—the very officials we presume reliable. As with a country report, the information in a consular letter may be based, in part, on hard-to-verify statements made by local officials. That’s a reason to take the information contained in such letters with a grain of salt—as an IJ is entitled to do—not a reason to deem them inadmissible in their entirety.

The similarities between the Bunton Letter and the litany of documents used, and accepted, in everyday asylum adjudications speaks to a fundamental misapprehension on the part of the Second Circuit and the dissent. Immigration adjudication necessarily requires consideration of all manner of imperfect sources. But we do neither immigrants nor the immigration authorities a service by cabining the range of permissible documents on which a trier of fact can rely in making his decision. In assessing whether an incident occurred years ago in a faraway country with an unfamiliar culture and political system, an immigration judge must be able to read, assess and weigh as much information as possible. True, dismissing a petition in reliance on an unsworn letter might seem harsh; but so is dismissing a petition based on relatively minor testimonial inconsistencies in the convoluted story of an immigrant who may have only a weak command of English and a hazy memory of his flight from terror. Harshness is endemic to any asylum system. Here, the IJ came to the conclusion that the unrefuted contents of the Bunton Letter cast doubt on the subpoenas Angov presented as evidence. Do we really better serve justice, or the immigration process more generally, by compelling the IJ to either accept the dubious subpoenas as genuine, or base his review solely on his instincts as to what a Bulgarian subpoena “should” look like?

An implicit assumption of the Second Circuit’s approach is that the exclusion of documents such as the Bunton Letter will lead, not to reliance on capricious information, but to the proliferation of more comprehensive and reliable State Department investigations. There’s no reason to believe that will happen. The asylum unit of the Department of State’s Office of Country Reports and Asylum Affairs “has suffered from long standing resource problems.” Office of the Inspector Gen., U.S. Dep’t of State, Report of Inspection: Bureau of Democracy, Human Rights and Labor 23 (2003). Many of its staffers are interns, and even its regular employees are often “pressed into service to work” on the Office’s other main responsibility: country reports. Id. at 23–24. And the consular officers tasked with verifying asylum applicants’ claims are also overworked and understaffed. The Tongour Letter expresses the

ANGOV V. HOLDER 25

government’s position on providing additional information about the results of an overseas investigation: “Such additional demands are further burdens on Consular Officers in the performance of their regular responsibilities and are particularly onerous for FSNs who may be subject to local reprisal.” The State Department tells us it’s doing the best it can with the scant resources allocated to it and our consular corps abroad.

Demanding that the reports contain a multitude of additional details, such as “the identity and qualifications of the investigator(s),” “the objective and extent of the investigation” and “the methods used to verify the information discovered,” see Lin, 459 F.3d at 271, transforms a process that is swift, efficient and informal into one that’s ponderous, time-consuming and expensive.

Insisting on these procedures would paralyze the process, making it impossible for our consular officers to do many of these investigations because they’re too busy filling in all the jots and tittles our sister circuit enshrines as pre-requisites for a document’s admission. Complying with such requirements considerably lengthens the time it takes to write most reports, and may make it impossible to write others for fear of disclosing sensitive information that could compromise sources or impair relations with local officials.

Nor is it realistic for the government to produce such information in camera. These reports are prepared by Department of State officials stationed in foreign countries, and are then turned over to another agency in another department, which then releases them to an adverse party. These disclosures are made in the context of immigration court proceedings, not in district court, and the immigration court, despite its name, is an executive branch agency. It has no contempt powers and can’t have anyone arrested for violating its orders, including confidentiality orders. See Stephen H. Legomsky, Restructuring Immigration Adjudication, 59 Duke L.J. 1635, 1674, 1714 (2010); Dana Leigh Marks, Still a Legal “Cinderella”? Why the Immigration Courts Remain an Ill-Treated Stepchild Today, 59 Fed. Law., Mar. 2012, 25, at 30. There’s a good chance the information will fall into the hands of people who have little regard for U.S. law and find themselves repatriated with a motive for revenge. Consular officials forced to disclose sensitive information in these circumstances would probably leave the information out of the report rather than risk burning their sources, offending local officials or losing their lives.

If we make the job of compiling these reports substantially more risky and onerous, the State Department may stop writing them. The United States gets close to 74,000 asylum cases a year, far more than any other industrialized nation. See United Nations High Comm’r for Refugees, Asylum Levels and Trends in Industrialized Countries 3, 8 & n.14 (2011). (That’s more than three times the number of Social Security cases the Supreme Court considered massive in Perales). The use of reports from consular officials gives the government the ability to check facts and puts at least some constraint on how far from the truth asylum applicants will stray. Knocking out even this most basic check on fraud and fabrication would subvert the asylum process, giving charlatans a free pass into the United States.

4. In any event, even if the Second Circuit’s approach were to encourage more detailed State Department letters,

ANGOV V. HOLDER 27

such “faith in procedural choreography” as a truth-seeking device is “fundamentally flawed.” United States v. Balough, 820 F.2d 1485, 1491 (9th Cir. 1987) (Kozinski, J., concurring). Requiring the Department of State to disclose more details will neither materially enhance the reliability of the resulting report nor do very much to help asylum applicants.

We test this proposition by modifying a portion of the Bunton Letter to comply with the requirements that would (presumably) satisfy the Second Circuit; new or modified language is italicized:

Agent Michael Smith, a foreign service agent with seventeen years of field experience who is fluent in Bulgarian, ordered Vladimir Popov, a foreign service national in the Embassy’s employ, to visit the 5th Police District station in Sofia in order to seek authentication of the two subpoenas. FSN Popov is a lifelong resident of Sofia and has worked for the Embassy for two years. He is fluent in Bulgarian and speaks conversational English.

FSN Popov traveled to the station and, once there, spoke to Ludmilla Bogdanovich, who is the supervisor of personnel records at the station. FSN Popov considers Ms. Bogdanovich a trustworthy source. After she consulted the relevant records, Ms. Bogdanovich told FSN Popov that Captain Donkov, Lieutenant Slavkov and Investigator Vutov have never worked for the 5th Police District. Ms. Bogdanovich also told FSN Popov that the case numbers on the subpoenas were not correct, there was no room 4 on the second floor and no room 5 on the first floor and that the telephone numbers on the subpoenas were incorrect. While at the station, FSN Popov asked Ms. Bogdanovich for an imprint of the police station seal, which he brought back to the consulate. Agent Smith compared it to the seal on the two subpoenas and found the official seal to be much larger.

After hearing FSN Popov’s oral report of his meeting with Ms. Bogdanovich, Agent Smith transmitted the information to the author of this letter by encrypted email.

Best we can tell, this revised letter would comply with the requirements imposed by the Second Circuit, but would it be much more valuable than what we already have? We’d know a bit more about Agent Smith, and we’d know the identity of the person who did the legwork, but how would that help us? We’d also have a name of someone who purportedly provided the information from the Bulgarians, but how would that be of any use? Angov could still complain that the IJ was unable to assess the Bulgarian official’s credibility, or even the credibility of any of the later links in the chain. We’d also know that it was Agent Smith who visually compared the seal on the subpoenas with the station’s official seal, but how does that bring us closer to the truth?

At this point, we would be faced with a whole new set of questions: How do we know Popov really went to the police

ANGOV V. HOLDER 29

station instead of stopping off in a bar to chug rakia? How did Popov know whether Bogdanovich was really the supervisor of personnel records at the police station? Did he check her identification papers? How did Popov assess Bogdanovich to be trustworthy, and how can we be sure he’s right? Did Popov look at the personnel records himself, or did he take Bogdanovich’s word that the three officers never worked there? Can we be sure that Bogdanovich checked all the relevant records? Can we be sure the purported personnel records were accurate and complete? How do we know Popov didn’t falsify important details because he was afraid of reprisal or because he hates gypsies? And how can we be sure Smith is telling the truth if we can’t cross-examine him? Did Smith have a full-sized copy of the subpoena when he compared the seals or a shrunken photocopy?

These difficulties are inherent in trying to prove up facts related to events that occurred years past and thousands of miles away from where the IJ is holding his hearing. Short of transporting all the declarants and their underlying records to the United States for a hearing before an IJ, there will inevitably be gaps that can be bridged only by multiple levels of hearsay.

This is not a problem that plagues only the government. Almost every piece of evidence asylum petitioners present in support of their cases would be inadmissible if subjected to the rules of evidence, especially those pertaining to hearsay: threats they claim to have been subjected to; racist comments by the police; reports of strange people looking for them; letters from family members and others. A brief scan of our caselaw shows it’s pretty much impossible to build an asylum case without relying on evidence that would be laughed out of court if presented in a domestic trial. See, e.g., Meza- Vallejos v. Holder, 669 F.3d 920, 922 (9th Cir. 2012); Haile v. Holder, 658 F.3d 1122, 1124–25 (9th Cir. 2011); Singh v. Holder, 656 F.3d 1047, 1049–50 (9th Cir. 2011); Hu v. Holder, 652 F.3d 1011, 1013–15 (9th Cir. 2011); Kumar v. Gonzales, 444 F.3d 1043, 1047–48 (9th Cir. 2006).

Take, as a small example, the letter from Daniela Mihaylova that Angov presented to rebut some of the information in the Bunton Letter. This is a two-page, typed document, with a small emblem and a typed address by way of letterhead. (We reproduce it in the Appendix.) It is addressed “To: Whom it may concern” and references Angov’s case. The letter represents that the “Romani Baht Foundation is a leading Bulgarian non-profit organization for protection of Roma/Gypsies human rights, founded in 1996 and legally registered with Bulgarian court.” Mihaylova purports to be the legal programs’ director of the Foundation.

The BIA took this letter seriously and modified some of the IJ’s findings based on it and other evidence presented by Angov. But there is absolutely no evidence in the record that there is any such person as Daniela Mihaylova and, if there is, how she went about obtaining the information detailed in her letter. For all we know, Angov could have printed the letter using his computer and standard word processing software.

Compared to this letter—and the remaining evidence presented by Angov—the Bunton Letter seems a paragon of reliability. It was prepared by government officials trained to perform this kind of investigation; who have nothing to gain by giving false information; and whose conduct is clothed with the presumption of regularity that attaches to all government actors. Cf. Perales, 402 U.S. at 402–06. The Bunton Letter encloses five photographs depicting locations

ANGOV V. HOLDER 31

mentioned in Angov’s asylum petition, which confirms that someone from our consulate traveled to those locations and made a personal inspection.

The Bunton Letter also gives specific reasons for doubting the authenticity of the addresses and points to several problems with the subpoenas. It is not an unsupported assertion that Angov is a liar; it is a rational, apparently objective recital of observed facts. At the very least, we can be sure that there is a Bunton and a Tongour, and that they can be disciplined or prosecuted if they negligently or deliberately falsified their reports. And we can reasonably presume that, in preparing their reports, Bunton and Tongour relied on trained State Department officers and agents who are themselves subject to discipline or prosecution for incompetence or corruption.

Compare this to the letter from Mihaylova (assuming there even is a Mihaylova): It comes from someone who cannot be disciplined or prosecuted in case of a lie, and who has not been screened for competence, honesty or reliability. It encloses no pictures or other documentary evidence. It doesn’t explain how the facts asserted were gathered or by whom. It doesn’t even claim to be based on first-hand knowledge, rather than hearsay or rumor. The letter simply makes a series of bald factual assertions without any support. Even assuming the letter is genuine (in the sense that it was actually written by its purported signatory in Bulgaria), the IJ and the BIA have absolutely no way to evaluate how accurate or objective it is.

In an environment where it’s pretty much impossible to obtain first-hand accounts of most of the relevant facts, should we require the government to fight an uphill battle on a slippery slope with one leg and both arms tied behind its back, while its adversary gets to use cleats and brass knuckles? Of course not. It would be the height of cognitive dissonance to hold the United States to standards of proof derived from domestic litigation while allowing petitioners to present anything and everything that doesn’t bear the watermark “Forgery Purchased on the Black Market.”

Furthermore, contrary to what the dissent and the Second Circuit might believe, the consequence of a rule excluding the consideration of documents such as the Bunton Letter will not be to allow more of the world’s oppressed into the land of the free. Rather, it favors the canny, the dishonest, the brazen and those who have the means and connections to purchase or create fraudulent documents, such as Angov’s compatriot, Pavlov. See p. 13–14 supra. Nor does such a rule ultimately help asylum seekers, as it’s hard to believe that Congress will long allow the program to continue when it rewards people who lie their way into the United States. Eventually, Congress and the public will catch on that asylum has become a fast-track vehicle for immigration fraud, and the asylum statute will be repealed or amended so as to make it even more difficult for honest asylum seekers to obtain relief. The ultimate victims will be the tired, poor, huddled masses who will find the golden door slammed in their faces.

* * *

We conclude on this record that the IJ acted within his discretion when he admitted the Bunton Letter into evidence and relied on it to find that the subpoenas Angov submitted were fraudulent. The adverse credibility finding based on the fraudulent subpoenas was supported by substantial evidence. Because Angov’s claim is based on his mistreatment by the

ANGOV V. HOLDER 33

Bulgarian police, the fact that the subpoenas were fraudulent “goes to the heart of [Angov’s] claim of persecution.” See Rizk, 629 F.3d at 1087–88. Furthermore, Angov’s testimony is not credible, and he doesn’t present other evidence that meets his burden to show that it’s “‘more likely than not’” that he would be tortured if sent back to Bulgaria. See Shrestha v. Holder, 590 F.3d 1034, 1048 (9th Cir. 2010). Consequently, the IJ and BIA decisions denying Angov asylum, withholding of removal and protection under the Convention Against Torture must stand.

PETITION DENIED. Appendix: Mihaylova Letter ANGOV V. HOLDER 35 THOMAS, Chief Judge, dissenting:

I would join the Second Circuit in resolving the issue before us. Unsworn, unauthenticated, hearsay letters— prepared for litigation by the government and not subject to any form of cross-examination—cannot form the sole basis for denying asylum to an otherwise qualified applicant. Therefore, I must respectfully dissent.

I

A

Five of our sister circuits have held that the government may not deny asylum solely on the basis of conclusory letters prepared for litigation in reliance on multiple layers of unauthenticated hearsay, without affording the petitioner some right of confronting the charges. Four of those circuits reached this result on constitutional grounds, holding that the admission of unauthenticated consular letters against an asylum applicant violates that applicant’s procedural due process rights. Banat v. Holder, 557 F.3d 886, 892–93 (8th Cir. 2009); Anim v. Mukasey, 535 F.3d 243, 256–58 (4th Cir. 2008); Alexandrov v. Gonzales, 442 F.3d 395, 407 (6th Cir. 2006); Ezeagwuna v. Ashcroft, 325 F.3d 396, 405–08 (3d Cir. 2003). The Second Circuit declined to reach the constitutional issue but held that such letters, standing alone, could not provide a basis for denying asylum under the substantial evidence standard because they lacked sufficient indicia of reliability and trustworthiness. Lin v. U.S. Dep’t of Justice, 459 F.3d 255, 268–72 (2d Cir. 2006); see also Balachova v. Mukasey, 547 F.3d 374, 382–83 (2d Cir. 2008) (applying Lin). Although I would resolve the present case purely on statutory grounds, as the Second Circuit did, the

ANGOV V. HOLDER 37

cases decided by our other sister circuits also provide useful guidance here.

In Banat, for instance, the Eighth Circuit rejected an IJ’s reliance on a consular letter that cited an unidentified investigator from the U.S. embassy in Beirut because the letter contained “multiple levels of hearsay” and omitted any mention of the investigator’s qualifications, experience, or “contact.” 557 F.3d at 891–92. The court acknowledged that “overseas investigations by State Department officials concerning the authenticity of documents purportedly originating in foreign countries are often necessary for the adjudication of an asylum claim,” id. at 890; however, it concluded that “the IJ’s reliance on the State Department letter, which provided no details about the investigation that would allow the IJ to assess the investigation’s reliability or trustworthiness and which contained multiple levels of hearsay, violated Banat’s right to a fundamentally fair hearing.” Id. at 893. The court reasoned:

Reliance on reports of investigations that do not provide sufficient information about how the investigation was conducted are fundamentally unfair because, without that information, it is nearly impossible for the immigration court to assess the report’s probative value and the asylum applicant is not allowed a meaningful opportunity to rebut the investigation’s allegations.

Id. at 891.

The Fourth Circuit relied on similar logic in Anim when it rejected a State Department letter authored by the same official involved in our case. The court noted that the official’s letter was “comprised entirely of multiple hearsay statements.” 535 F.3d at 257. It also pointed out that “letter does not explain how Bunton received the information she relates, nor does the letter disclose the identities of some of the individuals involved in the chain of communication.” Id.; see also id. (“Without the details of the investigation, it is impossible for an immigration judge, the BIA, or a court to evaluate the reliability of the letter’s conclusions.” (citations omitted)). Based on these deficiencies, the Anim court concluded that “the Bunton letter contains insufficient indicia of reliability and, as a result, its use was fundamentally unfair.” Id. at 256.

The courts in both Anim and Banat relied heavily on the Second Circuit’s reasoning in Lin, 459 F.3d at 268–72. In Lin, the Second Circuit rejected a consular report almost identical to the letter at issue here. The consular report was based on the opinions of Chinese government officials who, as the Lin court noted, “appear to have powerful incentives to be less than candid on the subject of their government’s persecution of political dissidents.” Id. at 269–70. The court also observed that the report lacked other traditional markers of reliability, namely: “(i) the identity and qualifications of the investigator(s); (ii) the objective and extent of the investigation; and (iii) the methods used to verify the information discovered.” Id. at 271. The Lin court distilled these factors from the Department of Justice’s own guidelines for evaluating the reliability of documents produced overseas.1 Noting that the consular letter failed to satisfy

1 The Justice Department’s guidelines stated that, in the case of a fraudulent document, the “report must contain, at a minimum: (i) the name and title of the investigator; (ii) a statement that the investigator is fluent

ANGOV V. HOLDER 39

these basic criteria, the court held that the report was “insufficiently detailed to permit a reviewing court to assess its reliability” and, as such, could not support a finding that the petitioner had forged documents submitted with his asylum application. Id. at 270.

Critically, the Second Circuit reached this conclusion as a statutory matter, holding that the consular report was “highly unreliable and therefore insufficient to satisfy the substantial evidence requirement.” Id. at 269. The court noted that the Third and Sixth Circuits had recently rejected similar reports as procedural due process violations2 but,

in the relevant language(s) or that he or she used a translator who is fluent in the relevant language(s); (iii) any other statements of the competency of the investigator and the translator deemed appropriate under the circumstances (such as education, years of experience in the field, familiarity with the geographic terrain, etc.); (iv) the specific objective of the investigation; (v) the location(s) of any conversations or other searches conducted; (vi) the name(s) and title(s) of the people spoken to in the course of the investigation; (vii) the method used to verify the information; (viii) the circumstances, content, and results of each relevant conversation or search[ ]; and (ix) a statement that the Service investigator is aware of the confidentiality provisions found in 8 C.F.R. § 208.6.” Memorandum from Bo Cooper (“Cooper Memo”), Gen. Counsel, Immigration & Naturalization Serv., to Jeffrey Weiss, Dir., Immigration & Naturalization Serv. Office of Int’l Affairs, Confidentiality of Asylum Applications and Overseas Verification of Documents and Application Information (June 21, 2001), available at http://judiciary.house.gov/ legacy/82238.pdf at 39–45. 2 See Alexandrov, 442 F.3d at 407 (holding that memoranda prepared by a U.S. embassy official in Sofia did “not meet our standards of trustworthiness and reliability and were therefore improperly relied upon by the immigration court”); Ezeagwuna, 325 F.3d at 406–08 (holding that a letter prepared by a U.S. embassy official in Yaounde contained “multiple hearsay of the most troubling kind” and, therefore, was “neither reliable nor trustworthy”). ultimately, the Lin court held that it was unnecessary to reach the constitutional issue. Id. (“Although we find the logic of these cases [concerning procedural due process] persuasive, we do not reach the constitutional issue presented because the statutory standard of review requires vacatur.” (emphasis in original)). The court later took the same approach in Balachova. 547 F.3d at 383 (concluding that a consular report that “contain[ed] no information concerning the qualifications of the investigators, the identity of the Russian officials who prepared the response to the consular inquiry, or the methods, if any, used to verify the information supplied by the foreign official” was “unreliable and cannot contribute to a finding of substantial evidence”).

Our case cannot be distinguished from Lin or Balachova. The IJ relied on a short, unsworn letter from a State Department official to support his finding that Angov forged parts of his asylum application. The letter was devoid of any information concerning the methodology employed in the investigation or the qualifications of the investigators. Instead, it was based on the unauthenticated, hearsay statements of an unidentified Bulgarian police official who worked at the police station where Angov claims to have been severely beaten. Like the government officials in Lin, that police official—whose department had been accused of ethnically motivated brutality—had a strong incentive to be “less than candid.” 459 F.3d at 269.

In sum, the Bunton Letter was comprised of conclusory statements of fact, none of which were supported by the basic information required under Lin and Balachova. We are left, as was the Second Circuit, with a document that is “insufficiently detailed to permit a reviewing court to assess its reliability.” Lin, 459 F.3d at 270. Indeed, in many ways,

ANGOV V. HOLDER 41

there is less information in the Bunton Letter than in the letters rejected as unreliable by our sister circuits. Accordingly, because the Bunton Letter lacks the indicia of reliability set forth in Lin, the agency could not have relied on it under the substantial evidence standard.

B

Neither Lin nor Balachova discussed the specific procedural protections guaranteed to aliens in removal proceedings under 8 U.S.C. § 1229a(b)(4)(B). That provision, however, offers independent grounds for barring the government from relying on unsworn, unauthenticated hearsay letters as the sole basis for denying an alien relief from removal.

Section 1229a(b)(4)(B) expressly provides that every alien “shall have a reasonable opportunity to examine the evidence against [him or her], to present evidence on [his or her] own behalf, and to cross-examine witnesses presented by the Government” during removal proceedings. See also 8 C.F.R. § 1240.10(a)(4) (requiring the IJ to “[a]dvise the respondent that he or she will have a reasonable opportunity to examine and object to the evidence against him or her, to present evidence in his or her own behalf and to cross-examine witnesses presented by the government”). We have recognized the “importance of the right to confront evidence and cross-examine witnesses” under this statute.3

3 The majority suggests that, because foreign officials are not themselves “amenable to cross-examination,” allowing State Department officials to be cross-examined when they “inescapably rely” on information from foreign officials would not significantly enhance the credibility of that information. Slip Op. at p. 11. This view overlooks the fact that State Cinapian v. Holder, 567 F.3d 1067, 1074 (9th Cir. 2009) (listing “several cases” highlighting the importance of this right). As we explained in those cases, the “purpose of this statutory guarantee cannot be fulfilled . . . if the government’s choice whether to produce a witness or to use a hearsay statement is wholly unfettered.” Baliza v. INS, 709 F.2d 1231, 1234 (9th Cir. 1983). Rather, to comply with this provision, the government must “make a reasonable effort to present the witness” for cross-examination. Cinapian, 567 F.3d at 1074.

The majority asserts that the “government here did make a reasonable effort to obtain a witness” for cross-examination but, ultimately, was stymied by the State Department’s policy of not releasing follow-up information about overseas investigations. See supra, Slip Op. at p. 11. This “effort” cannot be sufficient to satisfy the government’s burden under the statute. Indeed, allowing one executive branch agency to rely on another executive branch agency’s blanket policy of refusing to provide certain information is tantamount to granting the government the kind of unfettered discretion we repudiated in Baliza. As for the policy itself, whatever logistical obstacles might have once justified the State Department’s blanket refusal to produce overseas government witnesses for removal proceedings, those obstacles can surely

Department officials would be less likely to accept unreliable information as true if they knew that they might later be subject to cross-examination. Furthermore, if State Department officials did rely on information obtained from foreign officials, they would be prepared to explain why that information was trustworthy.

ANGOV V. HOLDER 43

be overcome in an age of video conferencing.4 Indeed, federal law specifically allows IJs to conduct entire hearings via telephone or video conference. 8 U.S.C. § 1229a(b)(2)(A); see also 8 C.F.R. § 1003.25(c) (“An Immigration Judge may conduct hearings through video conference to the same extent as he or she may conduct hearings in person.”).

Because the government did not make a reasonable effort to produce Bunton for cross-examination, I would hold that its reliance on the Bunton Letter violated Angov’s rights under § 1229a(b)(4)(B).5

C

The government argues that the Bunton Letter should be credited as trustworthy by employing the presumption of regularity—that is, that government officials accurately perform their reporting duties without bias. See Espinoza v.

4 At the very least, the State Department should be required to produce some specific hardship or reason why it cannot produce the witness for cross-examination, rather than relying on a general policy. 5 The four of our sister circuits to resolve this issue on constitutional grounds did not discuss the procedural rights guaranteed under § 1229a(b)(4)(B). However, the reasoning they used in concluding (unanimously) that the admission of unauthenticated, hearsay letters during removal proceedings violates an alien’s procedural due process rights counsels toward holding that such letters also violate the alien’s statutory rights under § 1229a(b)(4)(B). We have recognized that the due process right is closely related to the statutory right in this context. See Bondarenko v. Holder, 733 F.3d 899, 907 (9th Cir. 2013) (“The due process right, incorporated into 8 U.S.C. § 1229a(b)(4)(B), includes, among other things, ‘a reasonable opportunity to examine the evidence against the alien.’ (emphasis added; citations omitted)). INS, 45 F.3d 308, 310 (9th Cir. 1995) (holding that “information on an authenticated immigration form is presumed to be reliable in the absence of evidence to the contrary presented by the alien”).

However, the presumption of regularity does not apply when the source of information “was neither a government official nor the subject of the report.” Hernandez-Guadarrama v. Ashcroft, 394 F.3d 674, 681 n.9 (9th Cir. 2005) (citing Espinoza, 45 F.3d at 310). The key hearsay statement in the Bunton Letter comes from a Bulgarian police employee, not a U.S. government official or Angov. Statements made by third persons under no business duty to report are not entitled to the presumption of reliability and cannot be considered subject to the presumption, even if included in a document that enjoys such a presumption. United States v. Pazsint, 703 F.2d 420, 424-25 (9th Cir. 1983); see also Pouhova v. Holder, 726 F.3d 1007, 1014–15 (7th Cir. 2013) (rejecting application of presumption of reliability to hearsay statements of third parties recorded in official documents); Jordan v. Binns, 712 F.3d 1123, 1133 (7th Cir. 2013) (“[T]he presumption of reliability that serves as the premise for the public-records exception does not attach to third parties who themselves have no public duty to report.”).

Second, the presumption of reliability, similar to the traditional hearsay exception for public records, applies to documents “prepared in accordance with normal recordkeeping requirements.” Espinoza, 45 F.3d at 310; see also Lopez-Chavez v. INS, 259 F.3d 1176, 1181 (9th Cir. 2001) (“It must be shown that the document has been certified by the INS District Director as a true an[d] accurate reflection of INS records.”). The Bunton Letter, summarizing

ANGOV V. HOLDER 45

the results of an investigation involving multiple individuals and carried out at the behest of a party involved in litigation, is not comparable to an authenticated immigration form routinely filled out by border agents. Espinoza, 45 F.3d at 309. It is not a “business record” which is prepared in the usual and ordinary course of business. It was not authenticated or certified. It did not even conform with the agency’s own reporting procedures, as described and set forth in the Cooper Memo. Thus, the ad hoc Bunton Letter does not qualify as a government document produced in accordance with regular agency procedure.

For these reasons, I find the government’s arguments unpersuasive.

II

Adjudicating asylum claims is necessarily an imperfect endeavor. Witnesses to alleged foreign persecution are rarely available; documents are often impossible to locate. The immigration judge is often left with assessing witness credibility as the only means of resolving the request for relief. We are often limited to seeing through a glass, darkly.

As to post-REAL ID Act asylum seekers, the IJ may require corroboration, even when presented with credible testimony. See Aden v. Holder, 589 F.3d 1040, 1044 (9th Cir. 2009) (“Where the trier of fact determines that the applicant should provide evidence that corroborates otherwise credible testimony, such evidence must be provided unless the applicant does not have the evidence and cannot reasonably obtain the evidence.” (quoting 8 U.S.C. § 1158(b)(1)(B)(ii))). We have sustained the BIA’s denial of relief founded on the inability of an asylum seeker to obtain corroboration. Shrestha v. Holder, 590 F.3d 1034, 1047–48 (9th Cir. 2010).

In the post-REAL ID Act world, when corroborating evidence has assumed more importance, it is not unfair or unduly burdensome to require the government to identify basic, rudimentary information about its sources when it challenges corroborating evidence so that the IJ can properly weigh it. The information our sister circuits have demanded is modest. They do not require that every detail be uncovered or every riddle solved, they merely ask that very basic foundational questions—already in the hands of the Executive Branch—be answered. The Executive Branch invests significant resources in forensic document analysts, who provide detailed declarations in immigration cases. It is not much to ask that in the case of routine foreign fact- checking, the government simply tell us how it acquired the facts upon which it asks us to deny asylum.

The alternative is a decision founded solely on anonymous hearsay, often—as in this case—produced by the very foreign government actors the asylum-seeker accuses of persecution. We should be wary of relying on “secret informers, whisperers and talebearers” to decide legal rights in this context, especially when their word is used as the sole basis to deny relief to an otherwise qualified applicant. See Parker v. Lester, 227 F.2d 708, 719 (9th Cir. 1955) (cautioning against relying on untrustworthy sources in awarding security clearances to Coast Guard employees). The immigration system is fraught with enough risk of error. When it is reasonably possible, we need to minimize that risk.

I respectfully dissent.