join, concurring in part and dissenting in part;
I concur in Section II of the majority opinion, which overrules Molina-Camacho *1182v. Ashcroft, 393 F.3d 937 (9th Cir.2004). I respectfully dissent, however, from the majority’s outright denial of the petition for review on the merits. In this case, the immigration judge (“IJ”) granted asylum, but the Board of Immigration Appeals (“BIA”) reversed. In its reversal, the BIA made serious legal errors in its analysis and failed to fully address the claims made by the petitioner. Unlike the IJ, who relied on actual evidence, the BIA treated the aspirational goals of a government as a proxy for actual governmental control of those who would mercilessly persecute individuals based on their religion and national origin. Therefore, I would grant the petition and remand the case for further consideration by the BIA pursuant to INS v. Ventura, 537 U.S. 12, 123 S.Ct. 353, 154 L.Ed.2d 272 (2002).
I
The analysis used by the BIA was incorrect as a matter of law and cannot be sustained. The BIA’s analysis is legally flawed because it (1) failed to address a core element of the petitioner’s claim; (2) based its decision on grounds that are insufficient as a matter of law and in conflict with controlling case precedent; and (3) used the wrong standard of proof for asylum petitions. Each of these independent reasons requires a remand to the BIA for it to re-examine the case using the correct legal standards. See Ornelas-Chavez v. Gonzales, 458 F.3d 1052, 1058 (9th Cir.2006) (“[Wjhere the BIA applies the wrong legal standard to an applicant’s claim, the appropriate relief from this court is remand for reconsideration under the correct standard.... ”).
A
The BIA failed to address Lolong’s primary claim that there was a pattern or practice of persecution against Chinese Christian women that the Indonesian government was either unable or unwilling to control. All the BIA opinion decides is that the Indonesian government, in the BIA’s view, is willing to oppose discrimination and persecution of Christians and the ethnic Chinese, stating that the Indonesian government has “shown a general commitment to freedom of religion” and “a lack of institutional discrimination.” The BIA never discusses the Indonesian government’s ability to control the violence. Because there was substantial evidence of the government’s inability to control the violence regardless of its intentions, and because the IJ so found, the BIA’s reversal without explanation was error.
The BIA is obligated to consider and address in its entirety the evidence submitted by a petitioner. Mejia v. Ashcroft, 298 F.3d 873, 879-80 (9th Cir.2002); Kamalthas v. INS, 251 F.3d 1279, 1283-84 (9th Cir.2001). The BIA must “indicate with specificity that it heard and considered petitioner’s claims.” Maravilla Maravilla v. Ashcroft, 381 F.3d 855, 858 (9th Cir.2004) (quoting Arrozal v. INS, 159 F.3d 429, 433 (9th Cir.1998)). The BIA commits reversible error when it “merely repeat[s] petitioners’ claims and summarily dismisse[s] them without even purporting to engage in any substantive analysis or articulating any reasons for its decision.” Rodriguez-Lariz v. INS, 282 F.3d 1218, 1227 (9th Cir.2002).
Here, the IJ’s factual finding that the central Indonesian government has been unable to control the forces of persecution is amply supported by the voluminous record, but the BIA never addresses it. The record indicates that not only has the government been unable to control rogue forces in Indonesian society, but it has been unable to control its own military forces, which evidence suggests have carried out some of the grossest human rights abuses against Chinese and Christian minority groups.
*1183Since the 1998 riots and the ouster of President Suharto, the record shows that religious violence by Muslim fundamentalists against the Christian population has actually increased, often with the aid of military elements uncontrolled by the central government.1 Lolong presented substantial expert testimony on the issue, which the IJ credited and the BIA did not dispute. The expert testimony included an analysis of the current Indonesian regime, leading the expert to conclude that the present government could not control anti-Chinese and anti-Christian elements in Indonesia.
In short, the BIA did not address the pivotal point of Lolong’s claim — that the Indonesian government, for all its good intentions, is unable to control anti-Chinese and anti-Christian elements. There was substantial evidence in the record that such was the case and the IJ so found. Given the abundance of record evidence on the matter, the BIA was not permitted to avoid the question by claiming there was an “absence of evidence” that the government was unable to control the violence. See Ubau-Marenco v. INS, 67 F.3d 750, 760 (9th Cir.1995) (“the BIA must address the evidence presented, and state the reasons why ... [it] is not sufficient”) (internal quotation marks omitted), overruled on other grounds in Fisher v. INS, 79 F.3d 955, 963 (9th Cir.1996); Mejia-Carrillo v. INS, 656 F.2d 520, 522 (9th Cir.1981) (holding that the BIA must “give reasons which show that it has properly considered the facts which bear on its decision”). While the BIA was permitted to disagree with the IJ’s finding- — a decision we could have reviewed for substantial evidence — it was not permitted to ignore it altogether or reverse it without explanation.
This error was significant because if the BIA properly considered the evidence in full, I believe it should have concluded that the record amply supports the IJ’s conclusion that Ms. Lolong was objectively reasonable in her fear that she cannot rely on the government to protect her. A remand is required so that the BIA can address Lolong’s claim on the merits in the first instance.
B
The BIA’s decision was also insufficient as a matter of law and contrary to controlling case precedent. It determined that a Chinese or Christian could not reasonably fear persecution in Indonesia in early 2000 based on conclusions it drew from the State Department Country Reports. According to the BIA decision, the Reports show Indonesia’s “general commitment to freedom of religion and its lack of institutional discrimination against the ethnic Chinese minority.”
This rationale is insufficient as a matter of law. See Marcos v. Gonzales, 410 F.3d 1112, 1120-22 (9th Cir.2005). In Marcos, we specifically rejected the sufficiency of a Country Report statement much akin to *1184the one at bar, which stated that the government “generally respected” human rights. Id. at 1120 n. 7. Rather, we reiterated our long-standing rule that the agency use an individualized analysis with respect to the effect of State Department Country Reports.2 Id. at 1120-22. The BIA is not entitled to deny eligibility for asylum based on generalized conclusions in Country Reports. Lai v. INS, 255 F.3d 998, 1010 (9th Cir.2001).
Further, the BIA’s conclusion conflicts directly with Sael v. Ashcroft, 386 F.3d 922 (9th Cir.2004), in which we considered the very Country Reports at issue in this case. We detailed the “voluminous record” which documented “a history of anti-Chinese violence dating back as far as 1740.” Id. at 925. In addition to the 1998 riots in which more than 1,000 people were killed, we noted that “ethnic Chinese found their homes marked for destruction” in the years since the 1998 riots, and that “Chinese Christian churches and homes have ... been attacked in a violent merger of religious and ethnic tensions.” Id. at 925-26. We held that the very Country Reports at issue in this case buttressed the conclusion that “[g]overnment efforts to stop this cycle of scapegoating and violence have thus far been ineffective, perhaps because ethnic Chinese are still targets of official discrimination.” Id. at 926. We noted that the 1999 Country Report “ ‘found evidence that some elements of the military may have been involved in provoking the violence, which included attacks against Sino-Indonesian women’ during the 1998 riots.” Id. at 927. We found it significant that the 1999 Country Report stated that government regulations prohibit Chinese schools, cultural groups, and trade associations. Id.
We were not alone in reaching this conclusion. Former Chief Judge Becker wrote for the Third Circuit that:
In the late 1990s, Indonesia’s Chinese Christian population became the target of widespread attacks perpetrated by Muslim Indonesians. The 1999 United States State Department country report for Indonesia noted that “[i]nterreligious violence and violence against ethnic minorities continued. Attacks against houses of worship continued, and the lack of an effective government response to punish perpetrators and prevent further attacks led to allegations of official complicity in some incidents.” U.S. Dep’t of State, 1999 Country Reports on Human Rights Practices — Indonesia, February 25, 2000.[abbr. omitted]. In May 1998, there were “serious and widespread attacks” on Chinese-owned businesses and homes by Muslim Indonesians, which led to the deaths of over one thousand people.
Lie v. Ashcroft, 396 F.3d 530, 532-33 (3d Cir.2005). Thus, Judge Becker concluded that “1998 represented a period of significant violence and rioting against individuals of Chinese origin throughout Indonesia.” Id. at 533.
In short, the BIA was not permitted to rely on Country Reports to rebut Ms. Lo-long’s specific evidence about her family and her church and the general conclusions it drew from those Country Reports are precluded by case law.
C
The BIA applied the incorrect standard of proof in its analysis. The Supreme *1185Court has explicitly held that a petitioner seeking asylum is not required to prove that it would be more likely than not that she would be persecuted upon return to the country. INS v. Cardoza-Fonseca, 480 U.S. 421, 449, 107 S.Ct. 1207, 94 L.Ed.2d 484 (1987). Rather, under controlling precedent, a petitioner need only establish that there is a one in ten chance of persecution to show that the petitioner’s fear is objectively reasonable and that they are thus eligible for asylum. Njuguna v. Ashcroft, 374 F.3d 765, 770 (9th Cir.2004). The BIA applied the rejected “more likely than not” standard.3
The majority claims the BIA “clearly recognized” the difference between the two standards and applied them correctly, but a close look at the BIA’s opinion suggests otherwise. First, when the BIA introduced the “more likely than not” standard, it cited the refugee definition applicable to an asylum claim, I.N.A. § 101(a)(42)(A), and not the withholding of removal statute, § 241. Second, in applying the facts to the law on the asylum claim, it reasoned that Lolong’s fear was not well-founded because “the large majority of ethnic Chinese continue to reside in [Indonesia] without suffering physical attacks.”4 Yet if it was truly applying the correct one-in-ten standard, what happened to a “majority” of ethnic Chinese would be irrelevant. Indeed, that it considered the plight of the “majority,” or 51 percent, suggests it was applying the more likely than not standard, which by definition looks to the 51 percent mark. Third, whether Lolong was eligible for withholding of removal was not properly before the BIA. The IJ granted asylum relief, did not reach the question of whether Lolong was entitled to withholding of removal, and the INS only appealed the asylum holding.
The majority writes off any confusion by stating that “the BIA’s language could have been better crafted,” Maj. Op. at 1179 n. 3, but we have held that to avoid any confusion as to which standard the BIA is applying, “the BIA[is required] to make an explicit statement that it is applying a more generous standard to the asylum claim than to the petition for withholding of deportation.” Rodriguez v. INS, 841 F.2d 865, 869 (9th Cir.1987) (emphasis added). Not only is the BIA less than explicit about which standard it is applying, it never so much as mentions the proper one-in-ten standard.
The BIA’s legal analysis is impermissible under Cardoza-Fonseca, and this error of law requires remand for the BIA’s reconsideration “under the proper standard.” Martinez-Sanchez v. INS, 794 F.2d 1396, 1399 (9th Cir.1986).
II
In sum, I concur in the majority’s conclusion that Molina-Camacho should be overruled. However, I respectfully disagree with the majority as to the merits of Ms. Lolong’s claim. The BIA’s analysis on the merits was founded on an improper application of law in a number of respects. *1186It failed to address Lolong’s claims and the evidence supporting them, and reversed a core IJ finding without explanation. Its reasoning was insufficient as a matter of law and in conflict with controlling case precedent. It applied the incorrect standard of proof.
More disturbingly, the BIA rejected actual evidence in the record of the Indonesian government’s inability or unwillingness to control persecution in favor of the stated aspirational goals of that government. Remand is required so that the BIA can conduct a proper examination of the evidence under proper legal standards.
No one doubts that Ms. Lolong has a well-founded subjective fear of future persecution if she is forced to return to Indonesia. No one doubts her good faith, nor her contribution to our country during the time she has spent here. The only question in this case is whether her fear is reasonable. The undeniable fact is that over a thousand Chinese Indonesians were killed, churches were burned, and Chinese-Christian women raped just prior to the time when Lolong asked for asylum. By any objective measure, her specific fear of persecution was reasonable, and the BIA should have considered it on the merits of the evidence, under proper legal standards.
For these reasons, I respectfully dissent.
. For example, the record contains evidence that Indonesian Army units have been caught on tape "providing covering fire for Muslim gunmen attacking Christian neighborhoods.” Ron Moreau, An Island Holy War, Newsweek, August 7, 2000. The civilian Defense Minister reported that "members of the Army have become a major cause of the clashes,” and victimized Christians have reported seeing "military uniforms beneath the Jihad fighters' white robes.” Ian Timberlake, Indonesians see Suharto Behind Religious War; Survivors of Attack Blame Soldiers Loyal to exLeader, USA Today, August 2, 2000. "Instead of protecting them as promised, the Indonesian government soldiers joined in the slaughter” of Christians, witnesses say. Id. Members of the United States Congress have gone so far as to demand that the United States not lend any military expertise to Indonesian "military officers who approve of the killing of innocent women and children.” See USDS Pitts Decries Indonesian Military’s Role in Violence, Asia Pulse, July 26, 2000.
. In this context, it is also important to underscore that the BIA's analysis of country conditions did not come as part of "changed country conditions” examination, which occurs after a petitioner has established a prima facie case of a well-founded fear of persecution. It occurred as part of its initial examination of whether or not any reasonable person could fear persecution based on past events in the country.
. The “more likely than not” standard applies to the question of whether a petitioner is entitled to withholding of removal. 8 C.F.R. § 1208.16(b)(2); INS v. Stevie, 467 U.S. 407, 429-30, 104 S.Ct. 2489, 81 L.Ed.2d 321 (1984). Withholding of removal and asylum "are two distinct forms of relief.” Cardoza-Fonseca, 480 U.S. at 429 n. 6, 107 S.Ct. 1207. "[0]ur case law quite clearly establishes that the legal difference between ‘clear probability' and 'well-founded fear’ must be respected.” Rebollo-Jovel v. INS, 794 F.2d 441, 444 (9th Cir.1986).
. We know this discussion was part of the BIA’s asylum analysis because at the end of this discussion the opinion states "[n]or has the respondent established that ... she would qualify for withholding of removal.” (emphasis added).