IN RE R-A

FILPPU, Board Member:

In a decision dated September 20, 1996, an Immigration Judge granted the respondent's application for asylum under section 208(a) of the Immigration and Nationality Act, 8 U.S.C. § 1158(a) (1994). The Immigration and Naturalization Service has timely appealed the grant of asylum. The Service's request for oral argument before the Board has been withdrawn. The appeal will be sustained.

The question before us is whether the respondent qualifies as a "refugee" as a result of the heinous abuse she suffered and still fears from her husband in Guatemala. Specifically, we address whether the repeated spouse abuse inflicted on the respondent makes her eligible for asylum as an alien who has been persecuted on account of her membership in a particular social group or her political opinion. We find that the group identified by the Immigration Judge has not adequately been shown to be a "particular social group" for asylum purposes. We further find that the respondent has failed to show that her husband was motivated to harm her, even in part, because of her membership in a particular social group or because of an actual or imputed political opinion. Our review is de novo with regard to the issues on appeal. See Matter of Burbano, 20 IN Dec. 872 (BIA 1994).

A. Testimony and Statements of Abuse

The respondent is a native and citizen of Guatemala. She married at age 16. Her husband was then 21 years old. He currently resides in Guatemala, as do their two children. Immediately after their marriage, the respondent and her husband moved to Guatemala City. From the beginning of the marriage, her husband engaged in acts of physical and sexual abuse against the respondent. He was domineering and violent. The respondent testified that her husband "always mistreated me from the moment we were married, he was always . . . aggressive."

Her husband would insist that the respondent accompany him wherever he went, except when he was working. He escorted the respondent to her workplace, and he would often wait to direct her home. To scare her, he would tell the respondent stories of having killed babies and the elderly while he served in the army. Oftentimes, he would take the respondent to cantinas where he would become inebriated. When the respondent would complain about his drinking, her husband would yell at her. On one occasion, he grasped her hand to the point of pain and continued to drink until he passed out. When she left a cantina before him, he would strike her. As their marriage proceeded, the level and frequency of his rage increased concomitantly with the seeming senselessness and irrationality of his motives. He dislocated the respondent's jaw bone when her menstrual period was 15 days late. When she refused to abort her 3- to 4-month-old fetus, he kicked her violently in her spine. He would hit or kick the respondent "whenever he felt like it, wherever he happened to be: in the house, on the street, on the bus." The respondent stated that "[a]s time went on, he hit me for no reason at all."

The respondent's husband raped her repeatedly. He would beat her before and during the unwanted sex. When the respondent resisted, he would accuse her of seeing other men and threaten her with death. The rapes occurred "almost daily," and they caused her severe pain. He passed on a sexually transmitted disease to the respondent from his sexual relations outside their marriage. Once, he kicked the respondent in her genitalia, apparently for no reason, causing the respondent to bleed severely for 8 days. The respondent suffered the most severe pain when he forcefully sodomized her. When she protested, he responded, as he often did, "You're my woman, you do what I say."

The respondent ran away to her brother's and parents' homes, but her husband always found her. Around December 1994, the respondent attempted to flee with her children outside the city, but her husband found her again. He appeared at her door, drunk, and as she turned to leave, he struck her in the back of her head causing her to lose consciousness. When she awoke, he kicked her and dragged her by her hair into another room and beat her to unconsciousness.

After 2 months away, her husband pleaded for the respondent's return, and she agreed because her children were asking for him. One night, he woke the respondent, struck her face, whipped her with an electrical cord, pulled out a machete and threatened to deface her, to cut off her arms and legs, and to leave her in a wheelchair if she ever tried to leave him. He warned her that he would be able to find her wherever she was. The violence continued. When the respondent could not give 5,000 quetzales to him when he asked for it, he broke windows and a mirror with her head. Whenever he could not find something, he would grab her head and strike furniture with it. Once, he pistol-whipped her. When she asked for his motivation, he broke into a familiar refrain, "I can do it if I want to."

Once, her husband entered the kitchen where the respondent was and, for no apparent reason, threw a machete toward her hands, barely missing them. He would often come home late and drunk. When the respondent noted his tardiness, he punched her. Once, he asked where the respondent had been. When she responded that she had been home waiting for him, he became enraged, struck her face, grabbed her by her hair, and dragged her down the street. One night, the respondent attempted to commit suicide. Her husband told her, "If you want to die, go ahead. But from here, you are not going to leave."

When asked on cross-examination, the respondent at first indicated that she had no opinion of why her husband acted the way he did. She supposed, however, that it was because he had been mistreated when he was in the army and, as he had told her, he treated her the way he had been treated. The respondent believed he would abuse any woman who was his wife. She testified that he "was a repugnant man without any education," and that he saw her "as something that belonged to him and he could do anything he wanted" with her.

The respondent's pleas to Guatemalan police did not gain her protection. On three occasions, the police issued summons for her husband to appear, but he ignored them, and the police did not take further action. Twice, the respondent called the police, but they never responded. When the respondent appeared before a judge, he told her that he would not interfere in domestic disputes. Her husband told the respondent that, because of his former military service, calling the police would be futile as he was familiar with law enforcement officials. The respondent knew of no shelters or other organizations in Guatemala that could protect her. The abuse began "from the moment [they] were married," and continued until the respondent fled Guatemala in May 1995. One morning in May 1995, the respondent decided to leave permanently. With help, the respondent was able to flee Guatemala, and she arrived in Brownsville, Texas, 2 days later.

A witness, testifying for the respondent, stated that she learned through the respondent's sister that the respondent's husband was "going to hunt her down and kill her if she comes back to Guatemala."

We struggle to describe how deplorable we find the husband's conduct to have been.

Dr. Doris Bersing testified that spouse abuse is common in Latin American countries and that she was not aware of social or legal resources for battered women in Guatemala. Women in Guatemala, according to Dr. Bersing, have other problems related to general conditions in that country, and she suggested that such women could leave abusive partners but that they would face other problems such as poverty. Dr. Bersing further testified that the respondent was different from other battered women she had seen in that the respondent possessed an extraordinary fear of her husband and her abuse had been extremely severe.

Dr. Bersing noted that spouse abuse was a problem in many countries throughout the world, but she said it was a particular problem in Latin America, especially in Guatemala and Nicaragua. As we understand her testimony, its roots lie in such things as the Latin American patriarchal culture, the militaristic and violent nature of societies undergoing civil war, alcoholism, and sexual abuse in general. Nevertheless, she testified that husbands are supposed to honor, respect, and take care of their wives, and that spouse abuse is something that is present "underground" or "underneath in the culture." But if a woman chooses the wrong husband her options are few in countries such as Guatemala, which lack effective methods for dealing with the problem.

The Department of State issued an advisory opinion as to the respondent's asylum request. The opinion states that the respondent's alleged mistreatment could have occurred given its understanding of country conditions in Guatemala. The opinion further indicates:

[S]pousal abuse complaints by husbands have increased from 30 to 120 a month due to increased nationwide educational programs, which have encouraged women to seek assistance. Family court judges may issue injunctions against abusive spouses, which police are charged with enforcing. The [Human Rights Ombudsman] women's rights department and various non-governmental organizations provide medical and legal assistance.

The respondent has submitted numerous articles and reports regarding violence against women in Guatemala and other Latin American countries. One article, prepared by Canada's Immigration and Refugee Board, indicates that Guatemala has laws against domestic violence, that it has taken some additional steps recently to begin to address the problem, and that "functionaries" in the legal system tend to view domestic violence as a violation of women's rights. Nevertheless, the article indicates that Guatemalan society still tends to view domestic violence as a family matter, that women are often not aware of available legal avenues, and that the pursuit of legal remedies can often prove ineffective.

The Immigration Judge found the respondent to be credible, and she concluded that the respondent suffered harm that rose to the level of past persecution. The Immigration Judge also held that the Guatemalan Government was either unwilling or unable to control the respondent's husband. The balance of her decision addressed the issue of whether the respondent's harm was on account of a protected ground.

The Immigration Judge first concluded that the respondent was persecuted because of her membership in the particular social group of "Guatemalan women who have been involved intimately with Guatemalan male companions, who believe that women are to live under male domination." She found that such a group was cognizable and cohesive, as members shared the common and immutable characteristics of gender and the experience of having been intimately involved with a male companion who practices male domination through violence. The Immigration Judge then held that members of such a group are targeted for persecution by the men who seek to dominate and control them.

The Immigration Judge further found that, through the respondent's resistance to his acts of violence, her husband imputed to the respondent the political opinion that women should not be dominated by men, and he was motivated to commit the abuse because of the political opinion he believed her to hold.

On appeal, the Service argues that "Guatemalan women who have been involved intimately with Guatemalan male companions, who believe that women are to live under male domination" is not a particular social group, and that the respondent was not harmed because she belonged to such a group. The Service also contends that the respondent's husband did not persecute the respondent because of an imputed political opinion.

The respondent's brief supports the Immigration Judge's conclusions and advances additional arguments. The Refugee Law Center and the International Human Rights and Migration Project filed a joint amicus curiae brief. The thorough and well-prepared amicus brief argues that the Immigration Judge's decision is supported not only by United States asylum law, but also by international human rights laws, and that the respondent's asylum claims should be analyzed against the fundamental purpose of refugee law: to provide surrogate international protection when there is a fundamental breakdown in state protection resulting in serious human rights violations tied to civil and political status.

An asylum applicant bears the burden of proof and persuasion of showing that he or she is a refugee within the meaning of section 101(a)(42)(A) of the Act, 8 U.S.C. § 1101(a)(42)(A) (1994), to be eligible for asylum under section 208(a) of the Act. The term "refugee" refers to:

any person who is outside any country of such person's nationality . . . and who is unable or unwilling to return to, and is unable or unwilling to avail himself or herself of the protection of, that country because of persecution or a well-founded fear of persecution on account of race, religion, nationality, membership in a particular social group, or political opinion.

Section 101(a)(42)(A) of the Act;, 480 U.S. 421, 441 (1987).

We have held that members of a particular social group share a "common, immutable characteristic" that they either cannot change, or should not be required to change because such characteristic is fundamental to their individual identities. See Matter of Kasinga, 21 IN Dec. 357 (BIA 1996); Matter of H-, 21 IN Dec. 337 (BIA 1996); Matter of Acosta, 19 IN Dec. 211 (BIA 1985), modified on other grounds, Matter of Mogharrabi, 19 IN Dec. 439 (BIA 1987). The United States Court of Appeals for the Ninth Circuit, the circuit within which this case arises, defines a particular social group as:

a collection of people closely affiliated with each other, who are actuated by some common impulse or interest. Of central concern is the existence of a voluntary associational relationship among the purported members, which impart some common characteristic that is fundamental to their identity as a member of that discrete social group.

The asylum applicant bears the burden of providing evidence, either direct or circumstantial, from which it is reasonable to conclude that her persecutor harmed her at least in part because of a protected ground. See INS v. Elias-Zacarias, 502 U.S. 478, 483 (1992); Matter of T-M-B-, 21 IN Dec. 775 (BIA 1997), petition granted and remanded sub nom. Borja v. INS, 175 F.3d 732 (9th Cir. Apr. 30, 1999). The Court in Elias-Zacarias pointed out that overcoming or punishing a protected characteristic of the victim, and not the persecutor's own generalized goals, must be the motivation for the persecution. INS v. Elias-Zacarias, supra, at 482 ("The ordinary meaning of the phrase `persecution on account of . . . political opinion' . . . is persecution on account of the victim's political opinion, not the persecutor's.").

In rendering her decision, the Immigration Judge relied in part on the May 26, 1995, INS Asylum Gender Guidelines. See Phyllis Coven, U.S. Dep't of Justice, Considerations For Asylum Officers Adjudicating Asylum Claims From Women (1995) ("DOJ Guidelines"). On appeal, the parties argue over the applicability and force of the DOJ Guidelines to this case. We agree with the Immigration Judge that the guidelines make clear the point, which is independently evident on this record, that the level of harm experienced by the respondent rises beyond the threshold of that required for "persecution." The DOJ Guidelines also set forth various considerations for addressing "social group" and "political opinion" questions, but they provide no definitive answers for a case such as the one before us. Specifically, we do not read the DOJ Guidelines, which are instructive but not controlling on us, as resolving the issue of whether or when past spouse abuse may qualify a female applicant as a "refugee" under United States asylum law.

Similarly, we find no definitive answer in the language of the statute. Congress envisioned that the spouse of an alien granted asylum would ordinarily be accorded derivative asylee status, if he or she was not independently eligible. Congress provided for that derivative status, if the spouse were "accompanying, or following to join," the principal applicant. See section 208(c) of the Act (1994). Subsequent to enactment of the basic asylum provisions of current law in 1980, Congress has created specific forms of relief, outside our refugee laws, for some women living in or escaping from abusive marriages. See, e.g., section 240A(b)(2) of the Act, 8 U.S.C. § 1229b(b)(2) (Supp. II 1996) (cancellation of removal for spouses battered by a permanent resident or United States citizen); section 244(a)(3) of the Act, 8 U.S.C. § 1254(a)(3) (1994) (suspension of deportation for spouses battered by a permanent resident or United States citizen). No changes relative to battered spouses were made in the refugee definition or the asylum statute at the time of enactment of the battered spouse provisions.

The existence of derivative refugee status for spouses, as well as these nonrefugee provisions for battered spouses, raises the question whether Congress intended or expected that our immigration laws, even in the refugee and asylum context, would cover battered spouses who are leaving marriages to aliens having no ties to the United States. But we do not read the literal language of the statute actually to foreclose a construction that would accord refugee status to a battered spouse. In this case, we look principally to the facts to resolve both the "political opinion" and "social group" claims, and we do not intend any categorical rulings as to analogous social group claims arising under any other conceivable set of circumstances. Nevertheless, in reaching our decision, we find significant guidance in assessing the operation of the "particular social group" category by looking to the way in which the other grounds in the statute's "on account of" clause operate.

As noted above, we agree with the Immigration Judge that the severe injuries sustained by the respondent rise to the level of harm sufficient (and more than sufficient) to constitute "persecution." We also credit the respondent's testimony in general and specifically her account of being unsuccessful in obtaining meaningful assistance from the authorities in Guatemala. Accordingly, we find that she has adequately established on this record that she was unable to avail herself of the protection of the Government of Guatemala in connection with the abuse inflicted by her husband. The determinative issue, as correctly identified by the Immigration Judge, is whether the harm experienced by the respondent was, or in the future may be, inflicted "on account of a statutorily protected ground.

It is not possible to review this record without having great sympathy for the respondent and extreme contempt for the actions of her husband. The questions before us, however, are not whether some equitable or prosecutorial authority ought to be invoked to prevent the respondent's deportation to Guatemala. Indeed, the Service has adequate authority in the form of "deferred action" to accomplish that result if it deems it appropriate. Rather, the questions before us concern the respondent's eligibility for relief under our refugee and asylum laws. And, as explained below, we do not agree with the Immigration Judge that the respondent was harmed on account of either actual or imputed political opinion or membership in a particular social group.

The record indicates that the respondent's husband harmed the respondent regardless of what she actually believed or what he thought she believed. The respondent testified that the abuse began "from the moment [they] were married." Even after the respondent "learned through experience" to acquiesce to his demands, he still abused her. The abuse took place before she left him initially, and it continued after she returned to him. In fact, he said he "didn't care" what she did to escape because he would find her. He also hurt her before her first call to the police and after her last plea for help.

The respondent's account of what her husband told her may well reflect his own view of women and, in particular, his view of the respondent as his property to do with as he pleased. It does not, however, reflect that he had any understanding of the respondent's perspective or that he even cared what the respondent's perspective may have been. According to the respondent, he told her, "You're my woman and I can do whatever I want," and "You're my woman, you do what I say." In fact, she stated that "[a]s time went on, he hit me for no reason at all," and that he "would hit or kick me whenever he felt like it."

Nowhere in the record does the respondent recount her husband saying anything relating to what he thought her political views to be, or that the violence towards her was attributable to her actual or imputed beliefs. Moreover, this is not a case where there is meaningful evidence that this respondent held or evinced a political opinion, unless one assumes that the common human desire not to be harmed or abused is in itself a "political opinion." The record before us simply does not indicate that the harm arose in response to any objections made by the respondent to her husband's domination over her. Nor does it suggest that his abusive behavior was dependent in any way on the views held by the respondent. Indeed, his senseless actions started at the beginning of their marriage and continued whether or not the respondent acquiesced in his demands. The record reflects that, once having entered into this marriage, there was nothing the respondent could have done or thought that would have spared her (or indeed would have spared any other woman unfortunate enough to have married him) from the violence he inflicted.

Nonetheless, the Immigration Judge found support for her conclusion in the Ninth Circuit's decision in Lazo-Majano v. INS, 813 F.2d 1432 (9th Cir. 1987), overruled on other grounds by Fisher v. INS, 79 F.3d 955 (9th Cir. 1996). In Lazo-Majano, the alien's husband left El Salvador because guerrillas sought him and the government distrusted him for his involvement in and resignation from a paramilitary group. Id. A Salvadoran military sergeant who persecuted the alien denounced her as a "subversive," even though that accusation was done cynically as a means to control the alien better. The court ruled that the alien had suffered past persecution because of the subversive status attributed to her by the sergeant. Id. We understand this to be the holding of the case, and it is for this proposition that the case has continued to be cited within the Ninth Circuit. E.g., Meza-Manay v. INS, 139 F.3d 759 (9th Cir. 1998); Lopez-Galarza v. INS, 99 F.3d 954 (9th Cir. 1996).

We nevertheless recognize that the conduct of the sergeant toward the alien in Lazo-Majano paralleled in several respects the actions of the respondent's husband toward her here. In its decision, the court observed that "if the situation is seen in its social context," the sergeant was "asserting the political opinion that a man has a right to dominate and he has persecuted [the alien] to force her to accept this opinion without rebellion." Lazo-Majano v. INS, supra, at 1435. The court further suggested that the alien may be exposed to persecution from the sergeant because her flight could be seen as the expression of an opinion to the contrary. We have not, however, found a published Ninth Circuit case relying on this aspect of Lazo-Majano subsequent to the Supreme Court's ruling in INS v. Elias-Zacarias, supra, and we do not understand the "male domination" aspects of Lazo-Majano to be its actual holding.

Further, under Elias-Zacarias, the victim also must offer some evidence, direct or circumstantial, that it was the victim's political opinion that motivated the persecutor. INS v. Elias-Zacarias, supra, at 483. The respondent's husband, it seems, must have had some reason or reasons for treating the respondent as he did. And it is possible that his own view of men and women played a role in his brutality, as may have been the case with the brutality that he himself experienced and witnessed. What we find lacking in this respondent's showing, however, is any meaningful evidence that her husband's behavior was influenced at all by his perception of the respondent's opinion.

The respondent argues that, given the nature of domestic violence and sexual assaults, her husband necessarily imputed to her the view that she believed women should not be controlled and dominated by men. Even accepting the premise that he might have believed that the respondent disagreed with his views of women, it does not necessarily follow that he harmed the respondent because of those beliefs, rather than because of his own personal or psychological makeup coupled with his troubled perception of her actions at times. See id. at 482; Sangha v. INS, 103 F.3d 1482, 1487 (9th Cir. 1997) ("[T]he petitioner must prove something more than violence plus disparity of views.").

The Immigration Judge found, and the respondent argues, that her husband imputed a hostile opinion to her from her acts of resistance to his violence, and that he then punished her for that hostile opinion. The Court's ruling in Elias-Zacarias, however, establishes that the existence of a political opinion held by a persecutor, and actions by a victim that conflict with the demands of the persecutor, are not sufficient to require a conclusion that the persecutor seeks to harm the victim because of a contrary political opinion attributed to the victim. Both the respondent's argument and the "male domination" reasoning of Lazo-Majano seem to us to be akin to the analysis which the Supreme Court later did not accept as conclusive of political opinion persecution.

As we understand the respondent's rationale, it would seem that virtually any victim of repeated violence who offers some resistance could qualify for asylum, particularly where the government did not control the assailant. Under this approach, the perpetrator is presumed to impute to the victim a political opinion, in opposition to the perpetrator's authority, stemming simply from an act of resistance. Then, notwithstanding any other motivation for the original violence, the imputed political opinion becomes the assumed basis for the infliction of more harm.

It is certainly logical and only human to presume that no victim of violence desires to be such a victim and will resist in some manner. But it is another matter to presume that the perpetrator of the violence inflicts it because the perpetrator believes the victim opposes either the abuse or the authority of the abuser. We do not find that the second proposition necessarily follows from the first. Moreover, it seems to us that this approach ignores the question of what motivated the abuse at the outset, and it necessarily assumes that the original motivation is no longer the basis, at least not by itself, for the subsequent harm. We are unwilling to accept a string of presumptions or assumptions as a substitute for our own assessment of the evidence in this record, particularly when the reliability of these presumptions as genuine reflections of human behavior has not been established.

As for the record here, there has been no showing that the respondent's husband targeted any other women in Guatemala, even though we may reasonably presume that they, too, did not all share his view of male domination. The respondent was unable to set forth an accurate time frame for the great majority of the incidents she described. We are thus unable in general to link the incidents to acts of resistance in a way that might tend to support the respondent's theory. Moreover, the myriad situations in which the abuse occurred and the various unsuccessful responses adopted by the respondent point strongly away from it having a genesis in her husband's perception of the respondent's political opinion. Put another way, it is difficult to conclude on the actual record before us that there is any "opinion" the respondent could have held, or convinced her husband she held, that would have prevented the abuse she experienced.

Thus, unlike the aliens in cases such as Lazo-Majano, Lopez-Galarza, and Meza-Manay, the respondent here has failed to establish that her persecutor attributed to her a political view and then harmed her because of that view.

Initially, we find that "Guatemalan women who have been involved intimately with Guatemalan male companions, who believe that women are to live under male domination" is not a particular social group. Absent from this group's makeup is "a voluntary associational relationship" that is of "central concern" in the Ninth Circuit. Li v. INS, supra, at 987 (rejecting as a claimed social group Chinese citizens with low economic status); see also De Valle v. INS, supra, at 792 (rejecting as claimed social group "family members of deserters"); Sanchez-Trujillo v. INS, supra, at 1572 (rejecting as a claimed social group "young, working class males who have not served in the military of El Salvador").

Moreover, regardless of Ninth Circuit law, we find that the respondent's claimed social group fails under our own independent assessment of what constitutes a qualifying social group. We find it questionable that the social group adopted by the Immigration Judge appears to have been defined principally, if not exclusively, for purposes of this asylum case, and without regard to the question of whether anyone in Guatemala perceives this group to exist in any form whatsoever. The respondent fits within the proposed group. But the group is defined largely in the abstract. It seems to bear little or no relation to the way in which Guatemalans might identify subdivisions within their own society or otherwise might perceive individuals either to possess or to lack an important characteristic or trait. The proposed group may satisfy the basic requirement of containing an immutable or fundamental individual characteristic. But, for the group to be viable for asylum purposes, we believe there must also be some showing of how the characteristic is understood in the alien's society, such that we, in turn, may understand that the potential persecutors in fact see persons sharing the characteristic as warranting suppression or the infliction of harm.

Our administrative precedents do not require a voluntary associational relationship as a social group attribute. But we have ruled that the term "particular social group" is to be construed in keeping with the other four statutory characteristics that are the focus of persecution: race, religion, nationality, and political opinion. Matter of Acosta, supra. These other four characteristics are ones that typically separate various factions within countries. They frequently are recognized groupings in a particular society. The members of the group generally understand their own affiliation with the grouping, as do other persons in the particular society.

In the present case, the respondent has shown that women living with abusive partners face a variety of legal and practical problems in obtaining protection or in leaving the abusive relationship. But the respondent has not shown that "Guatemalan women who have been involved intimately with Guatemalan male companions, who believe that women are to live under male domination" is a group that is recognized and understood to be a societal faction, or is otherwise a recognized segment of the population, within Guatemala. The respondent has shown neither that the victims of spouse abuse view themselves as members of this group, nor, most importantly, that their male oppressors see their victimized companions as part of this group.

The lack of a showing in this respect makes it much less likely that we will recognize the alleged group as a particular social group for asylum purposes, or that the respondent will be able to establish that it was her group characteristic which motivated her abuser's actions. Indeed, if the alleged persecutor is not even aware of the group's existence, it becomes harder to understand how the persecutor may have been motivated by the victim's "membership" in the group to inflict the harm on the victim.

The respondent's showing fails in another respect, one that is noteworthy in terms of our ruling in Matter of Kasinga, supra. She has not shown that spouse abuse is itself an important societal attribute, or, in other words, that the characteristic of being abused is one that is important within Guatemalan society. The respondent has shown official tolerance of her husband's cruelty toward her. But, for "social group" purposes, she has not shown that women are expected by society to be abused, or that there are any adverse societal consequences to women or their husbands if the women are not abused. While not determinative, the prominence or importance of a characteristic within a society is another factor bearing on whether we will recognize that factor as part of a "particular social group" under our refugee provisions. If a characteristic is important in a given society, it is more likely that distinctions will be drawn within that society between those who share and those who do not share the characteristic.

Here, the respondent has proposed a social group definition that may amount to a legally crafted description of some attributes of her tragic personal circumstances. It may also be true that this description fits many other victims of spouse abuse.

In our opinion, however, the mere existence of shared descriptive characteristics is insufficient to qualify those possessing the common characteristics as members of a particular social group. The existence of shared attributes is certainly relevant, and indeed important, to a "social group" assessment. Our past case law points out the critical role that is played in "social group" analysis by common characteristics which potential persecutors identify as a basis for the infliction of harm. Matter of Kasinga, supra; Matter of H-, supra. But the social group concept would virtually swallow the entire refugee definition if common characteristics, coupled with a meaningful level of harm, were all that need be shown.

The starting point for "social group" analysis remains the existence of an immutable or fundamental individual characteristic in accordance with Matter of Acosta, supra. We never declared, however, that the starting point for assessing social group claims articulated in Acosta was also the ending point. The factors we look to in this case, beyond Acosta's "immutableness" test, are not prerequisites, and we do not rule out the use of additional considerations that may properly bear on whether a social group should be recognized in an individual case. But these factors are consistent with the operation of the other four grounds for asylum and are therefore appropriate, in our judgment, for consideration in the "particular social group" context.

On the record before us, we find that the respondent has not adequately established that we should recognize, under our law, the particular social group identified by the Immigration Judge.

Further, we cannot agree with the Immigration Judge's nexus analysis. In analyzing "particular social group" claims, our decisions, as well as those of the Ninth Circuit, in which this case arises, require that the persecution or well-founded fear of persecution be on account of, or, in other words, because of, the alien's membership in that particular social group. See Li v. INS, supra; De Valle v. INS, supra; Sanchez-Trujillo v. INS, supra; Matter of Kasinga, supra; Matter of H-, supra; Matter of Acosta, supra. This is reinforced by the Supreme Court's ruling in INS v. Elias-Zacarias, supra, at 483.

In this case, even if we were to accept as a particular social group "Guatemalan women who have been involved intimately with Guatemalan male companions, who believe that women are to live under male domination," the respondent has not established that her husband has targeted and harmed the respondent because he perceived her to be a member of this particular social group. The record indicates that he has targeted only the respondent. The respondent's husband has not shown an interest in any member of this group other than the respondent herself. The respondent fails to show how other members of the group may be at risk of harm from him. If group membership were the motivation behind his abuse, one would expect to see some evidence of it manifested in actions toward other members of the same group. See Li v. INS, supra (holding that even if Chinese citizens of low economic status did constitute a particular social group, the petitioner did not establish that authorities targeted members of that group); Sanchez-Trujillo v. INS, supra (finding that even if young, working class, urban males in El Salvador was a particular social group, the alien failed to demonstrate that the government singled out members of this group).

The Immigration Judge's nexus analysis fails to limit consistently the source of persecution to the respondent's husband. At one point, the Immigration Judge seems to identify all Guatemalan males who abuse their partners as the persecutors, but the record indicates that the respondent suffered and feared intimate violence only from her own husband. When the Immigration Judge correctly identifies the husband as the persecutor whom the Guatemalan Government failed to control, her nexus finding is both too broad and too narrow. It is too broad in that he did not target all (or indeed any other) Guatemalan women intimate with abusive Guatemalan men. It is too narrow in that the record strongly indicates that he would have abused any woman, regardless of nationality, to whom he was married.

Indeed, the record does not reflect that the respondent's husband bore any particular animosity toward women who were intimate with abusive partners, women who had previously suffered abuse, or women who happened to have been born in, or were actually living in, Guatemala. There is little doubt that the respondent's spouse believed that married women should be subservient to their own husbands. But beyond this, we have scant information on how he personally viewed other married women in Guatemala, let alone women in general. On the basis of this record, we perceive that the husband's focus was on the respondent because she was his wife, not because she was a member of some broader collection of women, however defined, whom he believed warranted the infliction of harm.

The respondent's statements regarding her husband's motivation also undercut the nexus claims. He harmed her, when he was drunk and when he was sober, for not getting an abortion, for his belief that she was seeing other men, for not having her family get money for him, for not being able to find something in the house, for leaving a cantina before him, for leaving him, for reasons related to his mistreatment in the army, and "for no reason at all." Of all these apparent reasons for abuse, none was "on account of a protected ground, and the arbitrary nature of the attacks further suggests it was not the respondent's claimed social group characteristics that he sought to overcome. The record indicates that there is nothing the respondent could have done to have satisfied her husband and prevented further abuse. Her own supposition is that he abused her because he was abused himself in the military.

The respondent was not at particular risk of abuse from her husband until she married him, at which point, given the nature of his focus, she was in a "group" by herself of women presently married to that particular man. Such a group, however, would fail to qualify as a "particular social group" under the Act. See Sanchez-Trujillo v. INS, supra; Matter of Acosta, supra.

The Immigration Judge nevertheless found, and the respondent argues on appeal, that her various possible group memberships account for her plight, in large measure because the social climate and the Government of Guatemala afford her no protection from her husband's abuse. Societal attitudes and the concomitant effectiveness (or lack thereof) of governmental intervention very well may have contributed to the ability of the respondent's husband to carry out his abusive actions over a period of many years. But this argument takes us away from looking at the motivation of the husband and focuses instead on the failure of the government to offer protection.

Focusing on societal attitudes and a particular government's response to the infliction of injury is frequently appropriate in the adjudication of asylum cases. It is most warranted when the harm is being inflicted by elements within the government or by private organizations that target minority factions within a society. But governmental inaction is not a reliable indicator of the motivations behind the actions of private parties. And this is not a case in which it has been shown that the Government of Guatemala encourages its male citizens to abuse its female citizens, nor in which the Government has suddenly and unreasonably withdrawn protection from a segment of the population in the expectation that a third party will inflict harm and thereby indirectly achieve a governmental objective.

The record in this case reflects that the views of society and of many governmental institutions in Guatemala can result in the tolerance of spouse abuse at levels we find appalling. But the record also shows that abusive marriages are not viewed as desirable, that spouse abuse is recognized as a problem, and that some measures have been pursued in an attempt to respond to this acknowledged problem. In this context, we are not convinced that the absence of an effective governmental reaction to the respondent's abuse translates into a finding that her husband inflicted the abuse because she was a member of a particular social group. The record does not support such a conclusion, as a matter of fact, when the husband's own behavior is examined. And Guatemala's societal and governmental attitudes and actions do not warrant our declaring this to be the case as a matter of law.

The Immigration Judge's decision relies heavily on the absence of governmental protection in its finding that the respondent was targeted for harm on account of her claimed group membership. The respondent takes this even further on appeal, arguing that governments can be deemed responsible for private acts of violence against women by virtue of the failure to afford protection. She also contends that she should be considered a "refugee" simply because she is not adequately protected by her own government.

We do not know whether enforcement measures would have deterred the abusive behavior of the respondent's husband in this case. But we do know that spouse abuse takes place even in communities with strong enforcement mechanisms. Varying levels of governmental tolerance of, or vigorous enforcement measures against, abuse can reasonably be expected to affect the incidence of spouse abuse within particular communities. It does not necessarily follow, however, that antagonism toward a "particular social group" is the motivation for the harm that husbands inflict upon their wives in those communities which afford little or no protection, even if certain societal attitudes may be seen as contributing to the absence of effective enforcement.

The adequacy of state protection is obviously an essential inquiry in asylum cases. But its bearing on the "on account of test for refugee status depends on the facts of the case and the context in which it arises. In this case, the independent actions of the respondent's husband may have been tolerated. But, as previously explained, this record does not show that his actions represent desired behavior within Guatemala or that the Guatemalan Government encourages domestic abuse.

Importantly, construing private acts of violence to be qualifying governmental persecution, by virtue of the inadequacy of protection, would obviate, perhaps entirely, the "on account of requirement in the statute. We understand the "on account of test to direct an inquiry into the motives of the entity actually inflicting the harm. See INS v. Elias-Zacarias, supra. Further, the adoption of such an approach would represent a fundamental change in the analysis of refugee claims. We see no principled basis for restricting such an approach to cases involving violence against women. The absence of adequate governmental protection, it would seem, should equally translate into refugee status for other categories of persons unable to protect themselves.

A focus on the adequacy of governmental protection would also shift the analysis in cases of refugee claims arising from civil war, as well as any other circumstance in which a government lacked the ability effectively to police all segments of society. This is not to say that the outcome of such an analysis would necessarily yield different results. The point, however, is that the existing statutory formula for assessing refugee claims would be altered. Instead of assessing the motivation of the actual persecutor, we might, for example, be focusing on the motivation or justification of the government for not intervening and affording real protection.

We reject the approach advocated by the respondent in view of the existing statutory language and the body of case law construing it. Consequently, the respondent must show more than a lack of protection or the existence of societal attitudes favoring male domination. She must make a showing from which it is reasonable to conclude that her husband was motivated to harm her, at least in part, by her asserted group membership.

In the end, we find that the respondent has failed to show a sufficient nexus between her husband's abuse of her and the particular social group the Immigration Judge announced, or any of the other proffered groups.

Our decision in Matter of Kasinga, supra, does not prescribe a different result. In that case, the alien belonged to the Tchamba-Kunsuntu tribe in Togo in which young women normally underwent female genital mutilation ("FGM") before the age of 15. Under tribal custom, the alien's aunt and husband planned to force her to submit to FGM before she was to be married. Following her escape from Togo, the Togolese police were looking for her. The record included a letter from a cultural anthropologist indicating that women from the Tchamba people probably would be expected to undergo FGM prior to marriage. A Department of State report in the record indicated that FGM was practiced by some Togo ethnic groups, that as many as 50% of Togolese females may have been mutilated, and that violence against women in Togo occurs with little police intervention. We held that FGM was persecution, that "young women of the Tchamba-Kunsuntu Tribe who have not had FGM, as practiced by that tribe, and who oppose the practice" constitute a particular social group, that the alien was a member of such a group, and that she possessed a well-founded fear of persecution on account of her membership in that group.

In contrast to our ruling in Matter of Kasinga, supra, the Immigration Judge in the instant case has not articulated a viable social group. The common characteristic of not having undergone FGM was one that was identified by Kasinga's tribe, and motivated both her family and the tribe to enforce the practice on Kasinga and other young women. Indeed, the tribe expected or required FGM of women prior to marriage, signifying the importance of the practice within that tribal society. The record in Kasinga indicated that African women faced threats or acts of violence or social ostracization for either refusing the practice or attempting to protect female children from FGM. Moreover, although the source of Kasinga's fear of physical harm was limited to her aunt and husband, she established that FGM was so pervasive that her tribal society targeted "young women of the Tchamba-Kunsuntu Tribe who have not had FGM, as practiced by that tribe, and who oppose the practice."

The respondent in this case has not demonstrated that domestic violence is as pervasive in Guatemala as FGM is among the Tchamba-Kunsuntu Tribe, or, more importantly, that domestic violence is a practice encouraged and viewed as societally important in Guatemala. She has not shown that women are expected to undergo abuse from their husbands, or that husbands who do not abuse their wives, or the nonabused wives themselves, face social ostracization or other threats to make them conform to a societal expectation of abuse. While the respondent here found no source of official protection in Guatemala, the young woman in Kasinga testified that the police in Togo were looking for her and would return her to her family to undergo FGM. Matter of Kasinga, supra, at 359.

We recognize that the respondent's situation is similar to that in Kasinga, in part, because the person actually inflicting the harm or feared harm is a family member of the victim. While the cases bear some similarities in this regard, we do not find this to be a factor that supports the claim of group recognition. Rather, it is a factor to be overcome if the group is to be accepted. In the context of asylum law, persecutors typically harbor animosities and act on those animosities toward many persons known to have the "hated" characteristic. When action is directed toward but one individual, or toward a small number of close family members, it calls into question both the propriety of the group definition and the alleged group motivation of the persecutor.

Furthermore, we may assume for purposes of discussion that "Guatemalan women who have been involved intimately with Guatemalan male companions, who believe that women are to live under male domination" constitute a particular social group. But, the respondent has not shown that Guatemalan males who believe in male domination and have been intimate with Guatemalan women actually target their intimate partners for persecution because of the victims' presumed group membership. The proposed social group represents a description of persons who may or may not experience harm. But it fails, on this record, as an adequate explanation for the surface, or even the more deep-rooted, factors that motivate the abusive behavior.

We find the dissent's analysis unconvincing, largely for the reasons we have already set forth in arriving at our decision. We shall thus confine our observations to the dissent's nexus arguments.

As the dissent correctly recognizes, we have granted relief in asylum cases where we have found it reasonable to believe that harm was inflicted, at least in part, because of a protected ground. E.g., Matter of S-P-, 21 IN Dec. 486 (BIA 1996). The dissent believes that the respondent's husband abused her, at least in part, because of an actual or imputed political opinion or because of her membership in a social group. It draws this conclusion from the cultural and societal context in which the abuse occurred, from literature indicating that domestic violence represents an exercise of power and domination over women, from her husband's view of her as his property, from the egregious nature of the harm, and from the absence of a legitimate motive for the abuse. We, on the other hand, do not find it reasonable to believe that an actual or imputed political opinion or social group membership led even in part to the respondent's abuse.

At the outset, the respondent never testified that she understood the abuse to be motivated by her political opinion or membership in a group of any description. Her husband never articulated such motivation, and she does not seem to have perceived it independent of the legal arguments now being advanced on her behalf. The dissent itself does not claim that either the respondent or her husband understood the abuse to be motivated, even in part, by the respondent's political opinion or social group membership.

In this context, the dissent's arguments for a political or social group motivation seem artificial. In our judgment, asylum law is not simply about the construction of various presumptions and inferences for bringing inarguably atrocious human action within one of the five grounds for which relief may be granted, particularly when those presumed or inferred motivations are undetected by both the abuser and the victim. For example, the perpetrators and victims of persecution because of race, religion, and political opinion typically understand and can explain the societal hatreds that lead to the harm or feared harm. We find it very difficult to accept the proposition that a persecutor targets persons who qualify as refugees for reasons that neither the persecutor nor the victims have been shown to understand as playing any role in the persecution.

In Matter of S-P-, supra, for example, we found that it was reasonable to believe that imputed political opinion played a role in the harm suffered by a person captured during a military operation and suspected of being a member of an armed opposition force in a civil war context. The political aspects of the conflict itself were readily apparent, and the participants on both sides well understood the conflict to have a significant political dimension. We were not required to presume the existence of a motivating factor that escaped recognition by any of the parties to the civil war. Our inquiry was simply to determine whether it was reasonable to believe that a known motivating factor in the existing conflict had actually contributed to the particular prisoner's torture.

In the case now before us, it simply has not been shown that political opinion or social group membership can reasonably be understood as the motivation behind the spouse abuse. Other factors, ranging from jealousy to growing frustration with his own life to simple unchecked violence tied to the inherent meanness of his personality, are among the explanations or motivations that may reasonably be inferred on this record for the actions of the respondent's husband. For example, when the respondent resisted her husband's demands for sexual relations, he would accuse her of seeing other men. Notably, he did not accuse her of harboring opinions hostile to his own or of being part of an abhorrent group.

The dissent also relies on the impunity with which the respondent's husband acted as support for its "on account of conclusions. In this regard, it draws on the opinion of Lord Hoffman in Islam (A.P.) v. Secretary of State for the Home Dep't, __ App. Cas. (March 25, 1999), available inhttp://www.parliament.the-stationery-office.co.uk/pa/ld9899/ ldjudgmt/jd990325/islamp1.htm, which argues that a Jewish businessman attacked by an Aryan competitor in Nazi Germany would be a victim of persecution on account of race because of the failure of the authorities to provide protection, even though the competitor was personally motivated only by business rivalry. But the very point of this example was to shift the focus away from the motivation of the entity causing the harm and to focus instead on governmental discrimination as satisfying the causation or nexus element for refugee status. Indeed, it does not appear that Lord Hoffman's nexus argument would be any different if the business competitor inflicting the harm had also been Jewish. The dissent's argument, consequently, is a variant of the respondent's claim that she should be accorded refugee status simply because she was not adequately protected by her government. We are not persuaded by this argument in the context of this case for the reasons we set forth earlier in addressing the respondent's contention.

We do agree with the dissent that the reasons set forth by the respondent's husband obviously do not in any way justify the abuse. But we find the lack of legitimate motives, an unconscionable level of harm, the escalation of the harm over time, and even the very incomprehensibleness of the abuse to be an inadequate basis from which to infer a statutorily qualifying motive. It is the respondent who bears the burden of proof. The dissent's approach, however, would seem effectively to shift the burden to the Service, as it would presume the existence of a qualifying case arising from serious harm and the absence of any apparently legitimate motive. We understand such an approach to free an asylum applicant of the need to offer evidence of motivation (as a qualifying motive would be presumed), and instead to force the Service to offer evidence of a "legitimate" reason for the infliction of the harm. In our judgment, it remains for the respondent to establish an evidentiary record from which we may reasonably infer that a qualifying motive led, at least in part, to the harm she suffered, and this she has failed to do.

In sum, we find that the respondent has been the victim of tragic and severe spouse abuse. We further find that her husband's motivation, to the extent it can be ascertained, has varied; some abuse occurred because of his warped perception of and reaction to her behavior, while some likely arose out of psychological disorder, pure meanness, or no apparent reason at all. Absent other evidence, we accept the respondent's own assessment that the foundations of the abuse she suffered lay in the abuse her husband had experienced in his own life. We are not persuaded that the abuse occurred because of her membership in a particular social group or because of an actual or imputed political opinion. We therefore do not find the respondent eligible for asylum, and consequently, she is ineligible for withholding of deportation under section 243(h) of the Act, 8 U.S.C. § 1253(h) (1994). See INS v. Cardoza-Fonseca, supra.

The respondent in this case has been terribly abused and has a genuine and reasonable fear of returning to Guatemala. Whether the district director may, at his discretion, grant the respondent relief upon humanitarian grounds — relief beyond the jurisdiction of the Immigration Judge and this Board — is a matter the parties can explore outside the present proceedings. We further note that Congress has legislated various forms of relief for abused spouses and children. The issue of whether our asylum laws (or some other legislative provision) should be amended to include additional protection for abused women, such as this respondent, is a matter to be addressed by Congress. In our judgment, however, Congress did not intend the "social group" category to be an all-encompassing residual category for persons facing genuine social ills that governments do not remedy. The solution to the respondent's plight does not lie in our asylum laws as they are currently formulated.

On the basis of the respondent's admissions, the Immigration Judge found the respondent deportable as an alien who entered without inspection. This finding is not challenged on appeal. In its September 1996, post-hearing brief before the Immigration Judge, the Service expressed its view that the respondent is statutorily eligible for voluntary departure and that it did not object to a grant of that relief. Accordingly, we will grant the respondent 30 days' voluntary departure. The respondent elected not to designate a country of deportation. Guatemala will be specified as the country of deportation pursuant to the Immigration Judge's determination at the December 7, 1995, hearing session.

Accordingly, the appeal will be sustained, and the respondent will be granted voluntary departure.

ORDER: The appeal of the Immigration and Naturalization Service is sustained and the Immigration Judge's order of September 20, 1996, is vacated.

FURTHER ORDER: In lieu of an order of deportation the respondent is allowed to depart voluntarily, without expense to the Government, within 30 days from the date of this order or any extension beyond that time as may be granted by the district director and under such conditions as he may direct. In the event of the respondent's failure so to depart, the respondent shall be deported to Guatemala.

Board Member Lori Scialabba did not participate in the decision in this case.

This Board acknowledges with appreciation the thoughtful arguments raised in amici curiae's brief.