On July 18, the Board of Immigration Appeals (BIA), the highest administrative immigration tribunal in the United States, issued a decision, Matter of K-E-S-G-, holding that a “particular social group” (PSG), a key term in the legal definition of a refugee, cannot be defined by sex (the term used in the decision in lieu of “gender”), whether alone or in combination with nationality.
The asylum seeker in this case, a young Salvadoran woman, was terrorized by members of the MS-13 intent on forcing her to submit to them sexually. She fled the country only after the police refused to respond to her pleas for help, and her efforts to shake off her tormentors by relocating twice were unsuccessful.
The ruling in K-E-S-G- is unpersuasive, yet it is also entirely consistent with a decades-long trajectory in BIA jurisprudence. As someone who has litigated these cases since the 1990s, I can say that for years, the Board has pursued two related objectives – rendering the PSG ground virtually impossible to satisfy by imposing ever-more rigid and illogical requirements, while ensuring an interpretation that would place most women fleeing gender violence outside the ambit of refugee protection. These goals have the full support of the Trump administration, which took aggressive steps during its first term to do the same. The administration’s position is especially significant because immigration and BIA judges are not independent decision-makers, as are Article III judges, but are employees of the Department of Justice, required to follow Attorney General rulings. In his second term, the Trump Justice Department has continued its restrictive rulings on social group and gender. Following shortly on the heels of the BIA’s decision in Matter of K-E-S-G-, Attorney General Bondi issued two additional rulings which continue this trend by limiting social group and gender claims; respectively, Matter of R-E-R-M & J-D-R-M- and Matter of S-S-F-M-.
Moving Goalposts of the “Particular Social Group” Definition
Under U.S. law, derived from the 1951 Refugee Convention and its 1967 Protocol, a refugee is an individual with a well-founded fear of persecution “on account of” race, religion, nationality, membership in a particular social group, or political opinion. The Convention drafters did not elaborate on the meaning of “particular social group,” but international refugee scholars have commented that its inclusion was to fill a possible gap in protection for claims not captured by the other four grounds, and subsequent case law in the United States and other jurisdictions has been consistent with that reading.
In 1985, the BIA had its first opportunity to interpret particular social group in Matter of Acosta. Noting the absence of interpretive guidance, the Board applied the doctrine of ejusdem generis which provides that a general term, included in a list of more specific terms, should be construed consistently with the specific terms. Reasoning that the other four Convention grounds refer to traits that a person cannot change, or should not be required to change to avoid persecution, the Board concluded that a “particular social group” is defined by immutable or fundamental characteristics, giving examples such as “sex, color, or kinship ties” or “shared past experience such as former military leadership or land ownership.”
Acosta’s immutable/fundamental approach, often referred to as the “protected characteristics” test, was cited approvingly by tribunals in other countries, including Canada, New Zealand, and the United Kingdom. At the same time as Acosta was decided, the U.N. High Commissioner for Refugees (UNHCR) began highlighting the relevance of the particular social group ground to claims by women seeking protection in its guidance. In three directives issued in 1985, 1991, and 1993, UNHCR recommended that although gender is not one of the five Convention grounds, women facing gender-based persecution could qualify as a particular social group and come within the refugee definition. UNHCR issued comprehensive guidance in affirmance of sex or gender based social groups in its 2002 Gender Guidelines. Although UNHCR guidance is not binding, the Board and other courts interpreting U.S. refugee law often look to UNHCR as persuasive authority, given that it provides authoritative interpretation on the international Convention from which U.S. law derives.
The BIA’s first published gender decision in 1996, Matter of Kasinga, was consistent with both Acosta and UNHCR guidance. It applied its immutable/fundamental analysis to grant protection to a woman fleeing female genital cutting, finding that she was a member of a social group defined by gender combined with other characteristics, including ethnicity and uncut status.
But only three years later, in Matter of R-A-, the BIA began to backtrack. It reversed a grant of asylum to a Guatemalan woman fleeing domestic violence, ruling that Acosta’s immutable/fundamental test was only a threshold, and that applicants must also show the proposed social group “was recognized and understood to be a societal faction,” which the BIA ruled, the applicant had failed to do. Attorney General Janet Reno vacated Matter of R-A-. She directed the BIA to decide the case once proposed regulations addressing gender claims which her Department of Justice had issued, were finalized – which, unfortunately, never occurred. (Notably, a February 2021 Executive Order committing to the promulgation of regulations on social group and gender claims by the Biden administration also never came to fruition.)
Beginning in 2006, the BIA formally ruled that a PSG could no longer be defined solely by immutable/fundamental characteristics. In a series of decisions, it added two additional requirements – particularity and social distinction. Particularity requires that a group be defined in a way that provides a “benchmark” for who is in, and who is out, while social distinction, previously referred to as “social visibility,” requires proof that the group is “perceived as a group by society.” In a previous article with Stephen Legomsky, we explained why we and other scholars disagree with the Board’s imposition of these two requirements.
Requiring proof of particularity and social distinction had a draconian impact on the success of social group claims, including those based on gender. The BIA’s only positive social group decision following the adoption of these criteria was Matter of A-R-C-G- in which the government stipulated to the cognizability of the social group.
The decision in A-R-C-G– was an early casualty of the first Trump administration. It was vacated by Attorney General Jeff Sessions, who wrote that “[h]ad the Board properly analyzed the issues,” rather than accepting the government’s stipulation, “it would have been clear that the particular social group was not cognizable.” During Biden’s term, Attorney General Merrick Garland reinstated A-R-C-G- as precedent, but Bondi vacated A-R-C-G- once again, by way of Matter of S-S-F-M- mentioned above.
What’s Next in K-E-S-G-?
The K-E-S-G- case was sent back to the immigration court for consideration of another form of relief, however its flawed reasoning will be challenged on appeal as it begins to be applied in other cases. In contrast to the BIA, a number of circuit courts have repeatedly recognized social groups based on gender, including groups defined solely by gender and nationality. In a series of decisions as early as 2005 and continuing to the present, courts have accepted gender plus nationality groups across a range of claims involving domestic violence, femicide, sexual assault, ritual harm inflicted on widows, and female genital cutting.
If upheld, the decision in K-E-S-G- would seriously undermine protection for women fleeing violations of their most fundamental human rights. But the weight of authority runs counter to its holding, and the decision’s rationales are particularly unconvincing.
The BIA disregards the growing body of contrary precedent, acknowledging only two of the many circuit court decisions that have recognized social groups defined by gender and nationality. It then attempts to discount one of those two by erroneously claiming that it was decided under a “prior definition” of social group.
The BIA leans heavily on the Third Circuit’s decision in Chavez-Chilel, which rejected the proposed group of “Guatemalan women.” But Chavez-Chilel is a confused and deeply flawed opinion. It misstates the governing law by holding that, for a group to be cognizable, every member must possess a well-founded fear of persecution. That has never been the standard. This reasoning improperly collapses two distinct elements of the refugee definition: the existence of a protected ground and the well-foundedness of the applicant’s fear. Numerous decisions have affirmed that each element must be considered separately.
The Board’s application of the social group requirements to the case before it – Salvadoran women – is perhaps the least persuasive aspect of the decision. While it concedes that the group’s defining characteristics are immutable, it rejects the claim on the basis of particularity and, having done so, declines to address social distinction. In effect, the entire rejection rests on its particularity analysis. The Board concludes that the group lacks particularity because it could encompass women of “varying ages, socioeconomic statuses, marital statuses, family backgrounds, and lifestyles.” Yet particularity requires only that a group be defined with sufficient clarity to establish its boundaries, and “Salvadoran women”—despite internal diversity—constitutes a group with clear and ascertainable limits. This is consistent with other findings of PSG in similar cases, such as the 1990 decision in Toboso-Alfonso, ruling that “Cuban homosexuals” is a cognizable group, which the BIA has more recently referenced as meeting the particularity and social distinction requirements.
In concluding—and perhaps in tacit recognition of the weakness of its legal analysis—the BIA turns to a quasi-policy justification for rejecting gender and nationality groups. It asserts that recognizing sex and nationality based social groups would amount to creating “another protected ground” under the statute, and that nothing in domestic law, the Refugee Convention and Protocol, or their negotiating history supports such an intent.
Whether the BIA genuinely believes that recognizing such groups would have this effect is debatable. In any event, the argument is peculiar. Allowing the social group ground to function as intended—filling gaps in protection—is not equivalent to adding a new protected ground. If it were, the recognition of every cognizable social group would constitute the creation of a new ground, a claim that has never been made.
Equally confounding is the BIA’s invocation of the Convention, the Protocol, and their negotiating history. UNHCR has for decades urged recognition of gender-defined social groups. Moreover, numerous State parties, through both legislation and jurisprudence, expressly recognize gender alone or gender-nationality groups.
The BIA’s efforts to justify its rejection of gender nationality social groups are ultimately unpersuasive. But K-E-S-G- will not be the final word. It is now up to the federal courts to address the issue. After the Supreme Court’s decision in Loper Bright, the federal courts are no longer bound by Chevron deference to “reasonable” agency interpretations of ambiguous statutory terms. One can only hope they reach the same conclusion as the First Circuit in de Pena-Paniagua.
[I]is not clear why a . . . group defined as “women” or “women in country x” – without reference to additional limiting terms – fails either the ‘particularity’ or “social distinction” requirement. . . . [I]t is difficult to think of a country in which women are not viewed as “distinct” from other members of society. . . It is equally difficult to think of a country in which women do not form a “particular” and “well-defined” group of persons.
In the meantime, the BIA is obligated to decide each case on its own record, so attorneys should continue to argue gender/nationality groups, along with narrower groups which may be more likely to be accepted.
