The Exception that Disproves the Rule: How Matter of K-E-S-G-‘s FGM Exception Exposes Its Incoherence
In its recent decision in Matter of K-E-S-G-, 29 I&N Dec. 245 (BIA 2025), the Board of Immigration Appeals (BIA) held that “a particular social group defined by the alien’s sex or sex and nationality, standing alone, is overbroad and insufficiently particular to be cognizable under the INA” as a basis for asylum. Matter of K-E-S-G-, 29 I&N Dec. at 152. Several organizations have already explained why this decision is an unlawful attack on refugee women that will have horrible consequences. In this blog post, rather than duplicating that work, I want to focus on why an exception that the BIA made in multiple footnotes of K-E-S-G- for cases relating to female genital mutilation (FGM) actually makes clear that the entire decision is logically incoherent.
The BIA in K-E-S-G- rejected the claim of the Salvadoran asylum applicant in that case, and seemingly attempted to pre-empt the asylum claims of all other women who assert that they face persecution due to their gender and nationality (absent other factors). In footnotes 7 and 8 of its decision, however, the BIA states that K-E-S-G- “does not involve a claim of female genital mutilation and our holding in this case does not affect the viability of such claims in the future” and on this basis distinguishes Hassan v. Gonzales, 484 F.3d 513 (8th Cir. 2007), which recognized a particular social group of Somali females in the context of a claim based on FGM.
The BIA’s exception for FGM was presumably made because of case law such as Hassan, and perhaps because of a more general realization that it would be facially ludicrous to deny that women subjected to FGM have faced persecution—although even if the BIA had taken that ludicrous step, Court of Appeals cases like Hassan and Mohammed v. Gonzales, 400 F.3d 385 (9th Cir. 2005), would exist whether the BIA acknowledged them or not. Upon further analysis, however, this exception for asylum claims based on FGM exposes why the overall holding of K-E-S-G- does not make sense.
The BIA says of Hassan that the Eighth Circuit there “held that “Somali females” was a particular social group because of the prevalence—98 percent—of female genital mutilation in the country.” Matter of K-E-S-G-, 29 I&N Dec. at 151 n.7. The implication seems to be that, according to the BIA, other forms of persecution of women are not so statistically prevalent, and that women in other contexts thus cannot constitute a particular social group even if they can constitute a particular social group in the FGM context in a country where the prevalence of FGM is so high. (The BIA does not make entirely clear how it would analyze an FGM case involving a lower prevalence than 98%, although footnote 8 of K-E-S-G- broadly exempts FGM claims from the decision’s holding without reference to a numerical cutoff.)
Asylum, however, does not require that persecution have a 98% likelihood, or anything close to that. The statutory standard under 8 U.S.C. § 1158(b)(1)(A) and 8 U.S.C. § 1101(a)(42)(A) is, instead, a well-founded fear of persecution. The Supreme Court clarified many years ago in INS v. Cardoza-Fonseca, 408 U.S. 421 (1987), that “to show a “well founded fear of persecution,” an alien need not prove that it is more likely than not that he or she will be persecuted in his or her home country.” Rather, “a showing of a ten percent likelihood of persecution could suffice to establish that an applicant’s fear is well-founded.” Kyaw Zwar Tun v. INS, 445 F.3d 554, 565 (2d Cir. 2006) (citing Cardoza-Fonseca, 408 U.S. at 431). Ordinarily, of course, one cannot quantify a likelihood of persecution with statistical certainty, but the basic point remains: the chance of persecution required in order to support an application for asylum is much, much lower than the 98% suggested by the evidence in Hassan.
Given this, however, the BIA’s basis for distinguishing Hassan and FGM while attempting to maintain an otherwise broad rule against gender as a particular social group is logically unsustainable. Under Cardoza-Fonseca, the outcome in Hassan should have been the same if only 10% of women in Somalia were subject to FGM, if that implied a 10% probability that a particular asylum applicant would be so subject by virtue of being a woman. The same is logically true of any other form of mistreatment, sufficiently severe to qualify as persecution, that a woman is exposed to with a 10% or greater likelihood, in a particular country, by virtue of being a woman—keeping in mind, again, that statistical certainty is usually not obtainable in these matters.
Since the regulations at 8 CFR 208.13(b)(1) and 8 CFR 1208.13(b)(1) provide that a victim of past persecution benefits from a presumption of a well-founded fear of future persecution, there will be many cases in which it will not make sense to require a (previously persecuted) asylum applicant to make any statistical showing at all. Moreover, a conflation of likelihood of persecution with the particularity of a particular social group is potentially problematic to begin with, since the evaluation of a particular social group and the evaluation of likelihood of persecution are supposed to be different stages of the asylum analysis. But even if we accept that an approximate statistical assessment may be relevant, as the BIA indicates in footnote 7 of K-E-S-G-, we must accompany that acceptance with the realization that under Cardoza-Fonseca, the relevant statistical threshold is nowhere near the 98% at issue in Hassan. If, as the BIA has implied in footnote 7, persecution of 98% of women in a particular country by means of FGM mandates their acceptance as a particular social group, then significantly lower rates of persecution of women ought to do so as well.
Even apart from the issue of the percentage chance of harm required to make out a claim, the BIA’s acknowledgement of an exception for the 98% prevalence of FGM in Somalia according to Hassan exposes another flaw in its logic. As the Court of Appeals for the Second Circuit explained in Ordonez Azmen v. Barr, 965 F.3d 128 (2d Cir. 2020), assessments of a particular social group must be done on a case-by-case basis, with reference to the record evidence pertaining to a particular country. The BIA has not overruled the case law relied upon by the Second Circuit in Ordonez Azmen (and perhaps could not do so without itself being overruled by a court), and yet it seems, in K-E-S-G-, to be trying to suggest that a social group of all women of a particular nationality can never be a particular social group, regardless of the record evidence regarding that particular country. In addition to this proposition being inconsistent with the Second Circuit’s decision in Ordonez Azmen (and thus legally non-viable within the Second Circuit or other Circuits that have held similarly), however, this proposition is falsified by the BIA’s own FGM exception. Apparently, the BIA accepts that the record in Hassan revealed such widespread persecution of women in Somalia that asylum was warranted in that case. But the BIA has no logical basis for categorically ruling out the possibility that the same could be true of some other country in which a sufficient number of women face some other form of persecution.
The BIA appears to have avoided immediate judicial review of its logically incoherent decision in K-E-S-G- by remanding the specific case to the Immigration Court for further consideration of an application for cancellation of removal, meaning that there is currently no final order of removal regarding which a petition for review could be filed under 8 U.S.C. § 1252. It is this author’s view, however, that lawyers handling other cases that present the possibility of a PSG based on gender should not merely accept the BIA’s decision in K-E-S-G- at face value, but should preserve the issue for future challenge on a petition for review, while of course also asserting any other PSGs or other protected grounds that remain available even under the BIA’s current view of the law. When the issue does come before the various Court of Appeals, some or all of those courts may recognize that the BIA’s attempt to cabin gender-based claims to one particular form of persecution, and an overly-demanding threshold of probability, should be rejected.