Lolong v. Gonzales

FOR PUBLICATION UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT MARJORIE KONDA LOLONG,  Petitioner, No. 03-72384 v.  Agency No. A77-427-355 ALBERTO GONZALES,* Attorney General, OPINION Respondent.  On Petition for Review of an Order of the Board of Immigration Appeals Argued and Submitted November 5, 2004—San Francisco, California Filed March 18, 2005 Before: Betty B. Fletcher, John T. Noonan, and Sidney R. Thomas, Circuit Judges. Opinion by Judge Betty Binns Fletcher *Alberto Gonzales is substituted for his predecessor, John Ashcroft, as Attorney General of the United States, pursuant to Fed. R. App. P. 43(c)(2). 3343 LOLONG v. GONZALES 3345 COUNSEL Hilari Allred, Law Office of Robert B. Jobe, San Francisco, California, for the petitioner-appellant. Lyle D. Jentzer, United States Department of Justice, Civil Division, Office of Immigration Litigation, for the respondent-appellee. 3346 LOLONG v. GONZALES OPINION B. FLETCHER, Circuit Judge: Marjorie Konda Lolong seeks asylum. Immigration Judge (“IJ”) Miriam Hayward granted relief, but the Board of Immi- gration Appeals (“BIA”) reversed. We grant the petition for review because compelling evidence establishes that Lolong has a well-founded fear of future persecution on account of her Chinese ethnicity were she returned to Indonesia. Specifi- cally, Lolong has demonstrated that Indonesians of Chinese ethnicity are a significantly disfavored group and that she is particularly at risk because she is also a Christian and a woman. I. Lolong has provided a voluminous record that documents ongoing discrimination and violence against the ethnic Chi- nese minority in Indonesia. We commented in detail on the long history of ethnic Chinese-Indonesian persecution in a similar case, Sael v. Ashcroft, 386 F.3d 922 (9th Cir. 2004). In Sael, we noted that there is a “cycle of waxing and waning violence” against ethnic Chinese-Indonesians. Id. at 929. Dur- ing periods of heightened social, economic, or political unrest, anti-Chinese sentiment erupts into wide-scale, severe vio- lence, but even during periods of “relative calm,” ethnic Chinese-Indonesians suffer discrimination and harassment, as well as violent attacks. Id. at 925-26, 929. When anti-Chinese violence erupted in May 1998, over a thousand ethnic Chinese-Indonesians were killed. Id. at 925-26. Although the Indonesian government has made some over- tures to the ethnic Chinese community, official anti-Chinese discrimination persists, and various forms of anti-Chinese vio- lence continue to occur. Numerous signs of economic, social, and political instability indicate that more severe anti-Chinese violence is likely to erupt again in the future. There is also LOLONG v. GONZALES 3347 evidence that rogue elements within the armed forces con- tinue to provide support to the nongovernmental forces that are responsible for ethnic and religious persecution. While all ethnic Chinese who remain in Indonesia face some risk of future persecution, two sub-groups are particu- larly at risk: ethnic Chinese women and Christians. Even dur- ing periods of relative calm, ethnic Chinese women can expect to be subjected to racial and sexual harassment when- ever they leave their homes. During the 1998 riots, at least dozens, possibly hundreds, of ethnic Chinese women were systematically raped. None of the perpetrators have been prosecuted. Should anti-Chinese sentiment erupt into more severe outbreaks of violence again, women will most likely be targeted again. Meanwhile, an Islamic fundamentalist move- ment continues to gain strength in Indonesia, and inter- religious conflict has been increasing. Not surprisingly, the combination of religious intolerance and ethnic prejudice has caused violent forces to target churches and homes in ethnic Chinese communities. Lolong left Indonesia after completing high school, because quotas limit the number of ethnic Chinese who are admitted to universities there. She was studying in the United States in May 1998, when the worst anti-Chinese rioting in Indonesia’s history occurred. After learning that one of her friends had been raped and her uncle had been severely beaten, Lolong decided to apply for asylum. In December 1998, Lolong filed a timely application. On November 16, 2000, after conducting an extensive hearing, Judge Hayward held that Lolong was eligible for asylum. She found Lolong to be fully credible and Lolong’s fear of future persecution on account of her ethnicity to be both subjectively and objectively reasonable. Subse- quently, in a divided opinion (2-1) the BIA sustained the Ser- vice’s appeal and vacated the petitioner’s application for asylum. Lolong timely petitioned for review. We have jurisdiction over Lolong’s petition pursuant to 8 U.S.C. § 1252(a). Because the BIA issued a reasoned opinion 3348 LOLONG v. GONZALES after conducting its own review of the record, we review the BIA’s decision for substantial evidence. Andriasian v. INS, 180 F.3d 1033, 1040 (9th Cir. 1999). In doing so, we accept Lolong’s testimony as true. Navas v. INS, 217 F.3d 646, 652 n.3 (9th Cir. 2000) (“Where the BIA does not make an explicit adverse credibility finding, we must assume that the applicant’s factual contentions are true.”). II. [1] In order to be eligible for asylum, Lolong must establish that she is a refugee—a person unable or unwilling to return to Indonesia “because of persecution or a well-founded fear of persecution on account of race, religion, nationality, mem- bership in a particular social group, or political opinion.” Sael, 386 F.3d at 924; 8 U.S.C. § 1101(a)(42)(A). The source of the persecution must be the government or forces that the govern- ment is unwilling or unable to control. Mashiri v. Ashcroft, 383 F.3d 1112, 1119 (9th Cir. 2004). To be “well-founded,” an asylum applicant’s “fear of per- secution must be both subjectively genuine and objectively reasonable.” Sael, 386 F.3d at 924. “An applicant satisfies the subjective component by credibly testifying that she genu- inely fears persecution.” Id. (internal quotation marks and citation omitted). Lolong satisfied this requirement with her credible testimony that she fears being hurt, raped, or killed in Indonesia. An asylum applicant “generally satisfies the objective com- ponent in one of two ways: either by establishing that she has suffered persecution in the past or by showing that she has a good reason to fear future persecution.” Id. While a well- founded fear must be objectively reasonable, it “does not require certainty of persecution or even a probability of perse- cution.” Hoxha v. Ashcroft, 319 F.3d 1179, 1184 (9th Cir. 2003). “Even a ten percent chance that the applicant will be persecuted in the future is enough to establish a well-founded LOLONG v. GONZALES 3349 fear.” Sael, 386 F.3d at 925 (quoting Knezevic v. Ashcroft, 367 F.3d 1206, 1212 (9th Cir. 2004)). [2] If not relying on a showing of past persecution, an asy- lum applicant’s fear of future persecution may be based on either individualized or group-based risk of persecution.1 See, e.g., Hoxha, 319 F.3d at 1182-83. Where an applicant estab- lishes that she is a member of a mistreated group, the level of individualized targeting that she must show is inversely related to the degree of persecution directed toward that group generally. Id. Accordingly, if an applicant demonstrates that she is a member of a group against which there is “a pattern or practice” of persecution, she need not further prove that she would be singled out for persecution. 8 C.F.R. § 208.13(b)(2)(iii); Kotasz v. INS, 31 F.3d 847, 852-53 (9th Cir. 1994). If the applicant demonstrates that she is a member of a “disfavored group,” but the group persecution does not rise to the level of a pattern or practice of persecution, then the applicant must also demonstrate that she is more likely to be targeted as a member of that group. Sael, 386 F.3d at 925 (collecting cases). A. [3] In Sael, we found that the long history of anti-Chinese discrimination, harassment, threats, and violence in Indonesia established that ethnic Chinese-Indonesians are “at least a dis- favored group.” 386 F.3d at 929. We stopped short of finding a “pattern or practice of persecution” against ethnic Chinese- Indonesians because of two statements in the record — “one that the Indonesian government officially promotes ethnic tol- erance and a second that racially motivated attacks against ethnic Chinese dropped sharply in 1999.” Id. (quotation marks omitted). We nonetheless concluded that these statements had 1 The IJ found that Lolong experienced harassment and discrimination but concluded that these past experiences did not amount to past persecu- tion. Lolong does not challenge this finding. 3350 LOLONG v. GONZALES little relevance to Sael’s asylum claim when compared to the voluminous evidence of continuing official discrimination and a centuries-old pattern in which anti-Chinese sentiment erupts into severe violence during periods of social, political or eco- nomic unrest. Id. Lolong has provided substantial evidence that the Indone- sian government, despite its official position on ethnic toler- ance, continues to discriminate against ethnic Chinese- Indonesians. For example, the Indonesian government contin- ues to enforce quotas limiting the number of ethnic Chinese that can be admitted into universities; Chinese culture remains excluded from school curriculums; and “[e]thnic Chinese con- tinue to face discrimination, particularly when they deal with the bureaucracy.” Although the government has begun turning a “blind eye” to the distribution of Chinese language media, the discriminatory laws prohibiting it remain in place. The record further establishes that anti-Chinese attacks and rioting are not only continuing to occur frequently but also are spreading to areas where anti-Chinese violence has been his- torically rare. Indeed, the Indonesian government’s decision to permit Chinese language press and cultural displays has caused heightened resentment and acts of retaliation. Dr. Syl- via Tiwon, a professor of Southeast Asian Studies at the Uni- versity of California at Berkeley whose research focuses on Indonesia,2 testified as an expert witness at Lolong’s hearing; she observed that there have been numerous attacks on Chi- nese and Christian cemeteries and explained that “retaliation against the . . . Chinese-Indonesian and the Christians hasn’t just taken place at . . . official and publicly seen levels.” Simi- larly, Lolong testified to receiving reports that a Chinese New Year celebration in Solo, Central Java, was attacked. At the time of Lolong’s immigration hearing, an ethnic Chinese 2 Dr. Tiwon has also conducted training sessions to acquaint asylum officers in the San Francisco office with conditions in Indonesia. LOLONG v. GONZALES 3351 human rights activist had been missing for two weeks; “disap- pearances” of activists are not uncommon in Indonesia. Even rioting that is not initially sparked by ethnic prejudice usually degenerates into anti-Chinese violence. For example, in May 2000, on the two-year anniversary of the infamous anti-Chinese riots that took place in May 1998, a clash between illegal vendors and police erupted into rioting; although it does not appear that anti-Chinese sentiment was the initial cause of the rioting, the mobs quickly turned on shops owned by ethnic Chinese. Ethnic Chinese women were attacked, ethnic Chinese-owned businesses were looted and destroyed. The expert witness explained that general rioting will continue to “turn into rioting against the Chinese” because they “have been the traditional scapegoats” and because the ethnic Chinese “are easily identified.” During the pendency of Lolong’s immigration hearing, both the Service and Lolong introduced numerous articles describing conditions in Indonesia. The expert witness was given the opportunity to review and comment upon these reports. She also provided information that she has gathered during her own research visits to Indonesia. She explained how the information in these reports, taken together, demon- strate that economic, social, and political conditions continue to be highly volatile. She also explained that continued insta- bility increases the probability that more severe outbreaks of anti-Chinese violence will erupt again in the future. She testi- fied that interviews she conducted with non-Chinese- Indonesians in 2000 reveal that ethnic Chinese-Indonesians continue to be scapegoated for the country’s widespread pov- erty and other economic problems. In Sael, we held that because ethnic Chinese-Indonesians are “significantly disfavored,” asylum applicants who prove that they are members of that group must demonstrate a “comparatively low” level of particularized risk. 386 F.3d at 927. At a minimum, Lolong’s new evidence of ongoing dis- 3352 LOLONG v. GONZALES crimination and violence, coupled with expert testimony sug- gesting that larger-scale outbreaks of violence are likely to occur again in the future, further lowers the level of particu- larized risk that Lolong must demonstrate in order to prove that her fear of future persecution is well-founded. Lolong’s evidence may even be strong enough to support a finding of a “pattern or practice” of persecution against ethnic Chinese in Indonesia within the meaning of 8 C.F.R. § 208.13(b) (2)(iii) (i.e. a pattern of persecution that negates the need for Lolong to make an additional showing of particularized risk).3 However, we need not decide whether the pattern of persecu- tion against ethnic Chinese-Indonesians rises to that level because the record compels us to conclude that Lolong has demonstrated sufficient particularized risk to support her asy- lum claim. [4] Although an asylum applicant may meet her particularized-risk burden by proving that she had past experi- ences that “establish a sufficient personal connection to the general persecution directed against” the disfavored group of which she is a member, Sael, 386 F.3d at 929, she may also meet this burden by proving she is a member of a sub-group that faces a heightened risk of future persecution. See Kotasz, 31 F.3d at 854 (explaining that the INS erred in construing the terms “individualized” and “singled out” literally). Even where members of a disfavored group[ ] are not threatened by systematic persecution of the group’s entire membership, the fact of group membership nonetheless places them at 3 Although the pattern or practice of persecution against ethnic Chinese- Indonesians does not appear to rise to the level of systematic ethnic cleansing, we have never held that only killings or extreme violence con- stitute “persecution.” Rather, we have repeatedly held that discrimination and harassment, in combination with lower levels of violence, may also constitute persecution. See, e.g., Korablina v. INS, 158 F.3d 1038, 1044 (9th Cir. 1998); Vallecillo-Castillo v. INS, 121 F.3d 1237, 1239 (9th Cir. 1996); Singh v. INS, 94 F.3d 1353, 1360 (9th Cir. 1996). LOLONG v. GONZALES 3353 some risk. That risk can rise to the level required for establishing a well-founded fear of persecution either as a result of an individual’s activities in support of the group, or because an individual is a member of a certain element of the group that is itself at greater risk of persecution than is the membership of the group as a whole. Id. at 853 (emphasis added). Moreover, “the more serious and widespread the threat to the group in general, the less individ- ualized the threat of persecution needs to be.” Sael, 386 F.3d at 925 (internal quotation marks and citations omitted). Lolong has met her burden by proving that she is not only a member of a significantly disfavored group but also a member of two sub-groups that are more likely targets for persecution: she is an ethnic Chinese Christian woman. The growth of a militant Islamic movement that has called for a jihad against Christian Indonesians4 — in a country where anti-Chinese prejudice is deep-seated and widespread and many ethnic Chinese are Christian — has led to a particu- larized pattern of violence against ethnic Chinese Christians. Although the anti-Chinese riots that occurred in May 1998 garnered much more international media attention, the State Department reported that there were outbreaks of “serious anti-Christian and anti-ethnic Chinese violence” as early as February 1998. United States Dep’t of State, 1999 Country Reports on Human Rights Practices: Indonesia [hereinafter “1999 Indonesia Country Report”]. Additional anti-Christian/ anti-ethnic Chinese riots occurred in the months of September and November, causing many deaths, injuries, and the destruction of churches, shops, and homes in predominantly ethnic Chinese communities. 4 Lolong provided both expert testimony and documentary evidence of a widespread militant Islamic movement that has been gaining strength in Indonesia, the leaders of which have been calling for a holy war and vio- lence against Christians. 3354 LOLONG v. GONZALES Apart from the violent outbreaks of rioting, attacks on eth- nic Chinese Christians and Christian houses of worship in eth- nic Chinese communities have become increasingly common. There have also been spates of church burnings coupled with burnings of Chinese residences. Demonstrations at Chinese Christian churches, during which violent threats are made against ethnic Chinese and Christian Indonesians, also occur frequently. Lolong’s parents decided to attend church less often after threats of violence were made during demonstra- tions near their church and after their church received bomb threats. The expert witness testified that ethnic Chinese- Indonesians’ risk of persecution increases with the rise of anti-Christian sentiment and violence. She also explained that the movement to “Islamisize business” is a movement to “take it out of the hands of the Chinese.” After describing out- breaks of anti-Christian and anti-Chinese violence, a State Department report similarly concluded that the link between anti-Christian and anti-Chinese violence is not coincidental: “Attacks on churches clearly reflect religious tensions, but other contributing factors are underlying socioeconomic and political tensions between poor Muslims and relatively more affluent ethnic Chinese Christians.” 1999 Indonesia Country Report. The record also establishes that the sub-group of ethnic Chinese Christian women is subject to a particularized risk of persecution. During periods of heightened ethnic and religious violence, there is a pattern of violence against women that rises to the level of persecution. During the May 1998 riots, dozens — possibly hundreds — of Chinese women were “systematically raped.”5 Mob clashes with police in Jakarta’s 5 It is well-established that rape and other forms of sexual or gender- based violence can constitute persecution on account of a protected ground. See, e.g., Shoafera v. INS, 228 F.3d 1070, 1075-76 (9th Cir. 2000) (holding where rape of asylum applicant was motivated at least in part by applicant’s ethnicity, applicant was persecuted “on account” of ethnicity as required for asylum eligibility). LOLONG v. GONZALES 3355 Chinatown, CNN.com, May 13, 2000; 1999 Indonesia Coun- try Report (noting that “[a]t least 85 instances of violence against women, including 66 rapes during the 1998 riots were verified” by a fact-finding team that included both govern- mental and nongovernmental investigators); Chris McCall, Ethnic Chinese Find a Voice, South China Morning Post, May 13, 2000 (reporting that a team of volunteers identified 163 ethnic Chinese rape victims from the 1998 riots and not- ing that the number of reported rapes may be low because volunteers found that women stopped coming forward after a victim who agreed to testify was murdered); Indonesian Chi- nese Call For Human Rights Investigation, World News Con- nection, May 12, 2000 (reporting that the Indonesian Ethnic Chinese Reforms Party collected data that suggested at least 300 ethnic Chinese women were raped during the 1998 riots); see also 1999 Indonesia Country Report (noting that violence against women and rapes are believed to be “seriously under- reported” in Indonesia). Two of Lolong’s friends from high- school were among the many ethnic Chinese women who were attacked. Although one friend managed to escape, the other was raped. Ever since, both young women have been afraid to leave their homes. Although the scale of the systematic raping of ethnic Chi- nese women during the 1998 riots is unmatched in Indonesian history, there is evidence that those attacks are reflective of a broader pattern in which Chinese and Christian minority women are singled out for abuse. For example, informal reports suggest that ethnic Chinese women were also the vic- tims of rape and/or attempted rape during the riot that took place in May 2000, on the two-year anniversary of the 1998 riots. Multiple reports confirm that both governmental and nongovernmental forces often use rape as a means of oppres- sion, but that the crime is rarely prosecuted. Even during peri- ods of relative calm, ethnic Chinese and Christian girls and women face a very high risk of being subjected to racial and sexual harassment and abuse, and ethnically and religiously motivated attacks against women continue to occur. The 3356 LOLONG v. GONZALES expert witness also testified that ethnic-Chinese women face a greater risk of being harassed or abused. The government argues that Lolong’s fear of future perse- cution is not reasonable because there is no evidence that Lolong’s “parents or brothers” have been harmed in any way in Indonesia on account of their ethnicity or religion. While the government correctly points out that the experiences of similarly situated friends or family who remain in the country are relevant to Lolong’s asylum claim, Hoxha, 319 F.3d at 1184, it conveniently ignores the fact that Lolong’s friends were attacked and raped, and continue to live in fear — afraid to even leave their homes. In other words, the experiences of persons most similarly situated to Lolong — two ethnic Chi- nese Christian women of Lolong’s age group and from the same community — support Lolong’s claim. That other mem- bers of Lolong’s family have not suffered from similar attacks is not surprising — Lolong is the only woman of the same age group in her family.6 Furthermore, contrary to the government’s assertion, many of Lolong’s family members have experienced at least some harm on account of their ethnicity or religion. Lolong’s father was arrested, detained for weeks at a time, and beaten, several times in the 1960’s, when the Indonesian government attempted to purge Indonesia of its ethnic Chinese minority by conducting “anti-communist” pogroms. More recently, several Indonesian youths attempted to rob Lolong’s uncle, but even after discovering that he was not carrying any money, they beat him so severely that he required surgery on his face. As noted above, the church to which Lolong’s par- ents belong has received bomb threats. As a result, her par- ents’ religious freedom has been curtailed. Although it is true 6 The “lack of evidence of harm” to one of Lolong’s brothers is irrele- vant, for he no longer lives in Indonesia. The other is currently unem- ployed; he believes that he cannot find employment, at least in part, because of ethnic discrimination. LOLONG v. GONZALES 3357 that Lolong’s mother has never been physically harmed, this is most likely due to the fact that she rarely leaves home, and only does so when accompanied. The loss of religious and personal freedom is a form of harm. Even if that harm does not rise to the level of persecution, it would be perverse to deny the reasonableness of Lolong’s fear on the ground that some of her family members have been able to avoid persecu- tion by giving up those fundamental freedoms. B. In rejecting Lolong’s claim of a well-founded fear of future persecution, the BIA concedes that discrimination and vio- lence targeting both ethnic Chinese and Christians continues to occur, yet, inexplicably, it insists that there is no evidence that the Indonesian government is unwilling or unable to con- trol the perpetrators. This conclusion is not supported by sub- stantial evidence. The substantial evidence of ongoing discrimination, harass- ment, and violence against ethnic Chinese-Indonesians could, without more, compel us to conclude that the Indonesian gov- ernment is “unable” to control the perpetrators. But Lolong did not rest her case on that evidence alone. The expert witness explained in detail why the Indonesian government’s authority is weak, and its ability to prevent eth- nic and religious violence is limited, despite its express com- mitment to promoting freedom of religion and ethnic tolerance. A State Department report similarly concludes that, notwithstanding the public positions of tolerance adopted by senior government officials, lower level officials frequently were alleged to be reluctant to facilitate and protect the rights of religious minori- ties. Minority houses of worship particularly have been targeted for damage or destruction during riots . . . Attacks against minority houses of worship and 3358 LOLONG v. GONZALES the lack of an effective government response to pun- ish perpetrators and prevent further attacks led to allegations of official complicity in some of the inci- dents or, at a minimum, allowing them to occur with impunity. United States Dep’t of State, Annual Report on Int’l Religious Freedom for 1999: Indonesia. There is substantial evidence that rogue elements within the Indonesian military are supporting the nongovernmental forces behind ethnic and religious persecution throughout the country. Various observers, including the State Department, have reported that the military has been supplying numerous unaccountable militias with arms and instigating political, religious, and ethnic violence. Government security forces may even be directly responsible for much of the violence. In 1999, the State Department reported that “[s]ecurity forces also were responsible for numerous instances of indiscrimi- nate shooting of civilians, torture, rape, beatings and other abuse” and that “[r]apes by security forces continued to be a widespread problem . . . .” 1999 Indonesia Country Report. During Lolong’s hearing, the Service made much of Presi- dent Wahid’s statements regarding his commitment to pro- moting ethnic tolerance, but ignored an August 2000 speech made by Wahid to the People’s Consultative Assembly, the highest legislative body in Indonesia, in which he conceded that the Indonesian economy had been “pulverized,” that security forces were unable to cope with violent religious con- flicts, and that the “rule of law was far from being estab- lished.” The Indonesian Minister of Defense has also announced publicly that “he cannot control various elements within the army.” There has been no prosecution — no arrests, no charges filed — in any of the rape crimes committed during the 1998 riots and their aftermath, even though a government inquiry LOLONG v. GONZALES 3359 confirmed reports by human rights groups that at least dozens of rapes were committed. Lolong testified that when one of her friends who was attacked during the riots reported the crime to the police, the police did nothing. Perpetrators reportedly intimidated surviving victims and witnesses. In the most extreme case, a rape victim who was willing to testify was murdered, and her body placed in the center of Jakarta. Supporters of the victims report that the police have done nothing to prevent or counteract this campaign of intimida- tion. Not surprisingly, the State Department has also found that the Indonesian government “did not resolve fully many cases of attacks on religious facilities and churches that occurred during riots and, in other cases, did not investigate such inci- dents at all.” 1999 Indonesia Country Report (internal citation omitted). The Indonesian government has not even followed the rec- ommendations of the fact-finding team that it commissioned to investigate the May 1998 riots. According to the State Department, the team “found evidence that some elements of the military may have been involved in provoking the vio- lence, which included attacks against Sino-Indonesian women, and urged further investigation in the matter.” 1999 Indonesia Country Report. Instead of prosecuting these ele- ments of the military, the People’s Consultative Assembly passed a resolution holding that “the army would not be held culpable for crimes against humanity committed before 1999.” There is also evidence that some military officers who are alleged to be responsible for instigating and orchestrating much of the violence are regaining power in Indonesia. [5] In its opinion, the BIA suggests that Lolong’s claim can, despite all of this evidence, be defeated by the mere fact that senior officials within the Indonesian government have publicly expressed their commitment to ethnic equality and religious freedom and instituted some reforms. However, this 3360 LOLONG v. GONZALES evidence of the government’s willingness to control the perpe- trators of ethnic and religious violence in Indonesia fails to rebut the overwhelming evidence of the government’s inabil- ity to control those forces. Cf. Avetova-Elisseva, 213 F.3d 1192, 1198 (9th Cir. 2000) (“It does not matter that financial considerations may account for [the government’s] inability to stop elements of ethnic persecution. What matters instead is that the government is unwilling or unable to control those elements of its society committing the acts of persecution.” (internal quotation and citation omitted) (emphasis in origi- nal)); Ladha v. INS, 215 F.3d 889, 894, 902 (9th Cir. 2000) (finding government “unwilling or unable to control” non- governmental forces where the government generally responded quickly to violence yet continued to experience a “serious law and order problem”). [6] In sum, Lolong has established by compelling evidence that she is a member of a significantly disfavored group, and that she is a member of sub-groups that are at a substantially greater risk of persecution than the group as a whole. As a result, she has met her burden of demonstrating that she faces a particularized risk of future persecution. The record also compels the conclusion that the Indonesian government is either unwilling or unable to control the forces behind ethnic and religious persecution in Indonesia. Accordingly, we find that Lolong’s fear of future persecution is well-founded. III. For the foregoing reasons, we grant the petition for review and hold that substantial evidence establishes Lolong’s eligi- bility for asylum. We remand so that the Attorney General may exercise his discretion to grant asylum.7 PETITION GRANTED; REMANDED. 7 We do not address withholding of removal or relief under the Conven- tion Against Torture because Lolong has not pursued those claims on appeal.