dissenting:
Because I find the majority opinion errs in holding that the Board of Immigration Appeals (“Board”) impermissibly interpreted the Immigration and Nationality Act (“INA”) in reviewing Temu’s application for asylum and withholding of removal under the INA, I respectfully dissent. Specifically, I conclude that the Board did not err as a matter of law in its determination that Temu’s proposed social group lacked the necessary characteristic of particularity.
I.
As the majority describes, the facts are not in dispute and the immigration judge (“U”) found the evidence Temu presented — including his own testimony about his treatment in Tanzania — to be credible. The record demonstrates that, beginning in 2001 or 2002, Temu began displaying erratic behavior and was both hospitalized and jailed following episodes where he would walk onto a road and attempt to direct traffic. During his hospitalizations and imprisonment, he was tied up and beaten. The nurses and doctors indicated that they believed Temu was mentally ill *898and referred to him by the term “mwenda wazimu,” which roughly means “demon-possessed or deranged.” (A.R.55, 136.1) Upon Temu’s release, his family increasingly rejected him as their efforts to rid him of his “evil spirit” failed. (A.R.56, 137.)
While lawfully in the United States on a temporary visa, Temu was taken to the hospital after police observed him standing in the middle of a street attempting to direct traffic. Over time and after observation, medical professionals diagnosed Temu with bipolar disorder. Temu’s visitor’s visa expired in 2006, but he nevertheless remained in the United States unlawfully. Although he can be employed and function independently while on medication, Temu has suffered relapses when he does not take his medications. On two occasions during such relapses, Temu was detained in psychiatric facilities after police observed him in the middle of a street attempting to control traffic.
In 2010, the Department of Homeland Security filed a notice to appear against Temu charging him with eligibility for removal based on overstaying his visitor’s visa without authorization. He asserted eligibility for asylum and withholding of removal under the INA based on his membership in a particular social group, which he identified as “schizophrenic and bipolar individuals in Tanzania who exhibit outwardly erratic behavior.”2 (A.R.72, 153.)
The IJ concluded that Temu satisfied the requirements for filing a late application for relief, and also determined that he satisfied the criteria for relief under the Convention Against Torture. As to Temu’s other claims, the IJ concluded that Temu was ineligible for asylum or withholding of removal under the INA because he failed to “demonstrate! ] that the harm he suffered was inflicted on account of his membership in a cognizable particular social group as required by law.” (A.R.72, 153.) She held that Temu’s proposed group lacked the three characteristics for a particular social group that the agency has determined the INA requires: particularity, immutability, and social visibility. Alternatively, the IJ concluded that Temu failed to demonstrate a nexus between his past persecution and his membership in the proposed social group. The Board adopted the IJ’s findings and conclusions, and dismissed Temu’s appeal, agreeing that Temu’s proposed group did not satisfy the requirements for a particular social group.
II.
The Court must uphold the denial of Temu’s application for asylum unless the denial is “manifestly contrary to the law and an abuse of discretion.” 8 U.S.C. § 1252(b)(4)(D). Whether a proposed group constitutes a particular social group under the INA is a question of law that the Court reviews de novo. Lin v. Mukasey, 517 F.3d 685, 691 (4th Cir.2008). In so doing, however, the Court accords Chevron3 deference to the Board’s reasonable interpretation of the INA. Id. at 691— 92. Where, as here, the Board adopts the IJ’s decision and includes its own reasons for affirming, the Court reviews both deci*899sions. Marynenka v. Holder, 592 F.3d 594, 600 (4th Cir.2010).
To establish eligibility for the discretionary relief of asylum, Temu had the burden of showing that he has a “well-founded fear of persecution on account of ... membership in a particular social group____” Naizgi v. Gonzales, 455 F.3d 484, 486 (4th Cir.2006) (citing 8 U.S.C. § 1101(a)(42)(A)); 8 C.F.R. § 1208.13(a). Temu faces a higher burden of proof to establish eligibility for withholding of removal — he must demonstrate “that it is more likely than not that [his] life or freedom would be threatened ... because of [his] ... membership in a particular social group.... ” Gomis v. Holder, 571 F.3d 353, 359 (4th Cir.2009) (citing 8 U.S.C. § 1231(b)(3)(A); 8 C.F.R. § 208.16(b)(2)); see also Camara v. Ashcroft, 378 F.3d 361, 367 (4th Cir.2004).
The INA does not define “particular social group.” The IJ and Board rely on prior Board opinions that interpreted this phrase to mean a group that (1) “share[s] a common, immutable characteristic” other than the fact that they are targeted for persecution, (2) that “[has] particular and well-defined boundaries,” and (3) that “possesses] a recognized level of social visibility.”4 (A.R.68.)
Temu challenges the IJ and Board’s application of the three-part analysis of his proposed group and its conclusion that it was not a cognizable “particular social group” under the INA. As noted, the IJ and Board concluded that Temu’s proposed group of bipolar individuals in Tanzania who exhibit outwardly erratic behavior lacked all three characteristics necessary for an alien to prove his or her “particular social group” status: immutability, particularity, and social visibility. I conclude that the Board’s decision can be upheld with respect to the particularity requirement, and it is therefore unnecessary to proceed further to deny Temu’s petition for review. See Zelaya v. Holder, 668 F.3d 159, 167 (4th Cir.2012) (affirming the Board’s decision to deny asylum based on conclusion that the proposed group lacks particularity -without discussing the remaining characteristics)
As its title suggests, a “particular social group” must have “particular and well-defined boundaries,” id. at 166, such that it is not “too amorphous ... to create a benchmark for determining group membership,” Matter of S-E-G-, 24 I. & N. Dec. 579, 584 (B.I.A.2008). The majority concludes that the IJ and Board’s determination that Temu’s proposed group lacks *900particularity is based on an error of law. Maj. Op. at 894-96. Specifically, the majority opines that the IJ and Board “commit[ted] legal error by splitting Mr. Temu’s group in two and rejecting each part, rather than considering it as a whole.” Maj. Op. at 895. I disagree.
At the outset, it is fundamentally important to review the IJ and Board’s actual holdings. The IJ concluded that Temu’s proposed social group lacked particularity “because the term ‘erratic behavior’ is too amorphous to ‘provide an adequate benchmark for determining group membership.’ ” (A.R.72.) It explained that “[e]r-ratic behavior could range from eccentric remarks to violent outbursts. Whether one perceives a particular behavior to be ‘erratic’ is a question of perspective, and as a result the respondent’s particular social group lacks particularity.” (Id.) The Board, in turn, agreed with the IJ’s conclusion that “ ‘erratic behavior’ is too indeterminate to meet the requirement of particularity ... as the mentally ill ... comprise a wide variety of different individuals covering a broad range of functionality.” (A.R.4.) And in rejecting Temu’s argument that he satisfied the particularity requirement “with evidence that individuals with ... bipolar disorder suffer from a discrete set of symptoms[] that include[s] erratic behavior,” the Board also noted that “erratic behavior is an amorphous characteristic, which does not provide an adequate benchmark for determining group membership or a concrete trait that would readily identify a person as possessing such a characteristic.” (Id.)
As the above recitation suggests, the Board did not divide Temu’s proposed group into two discrete subsections and consider them entirely apart from each other. To the extent that the underlying analysis focused on the component parts, however, it was not an error of law to do so. If each component of a proposed social group is amorphous and lacks particularity then it will often be true — and I conclude is the case here — that the group as a whole also lacks the requisite particularity. Far from being an aberration in this Court’s or the Board’s precedent, the IJ and Board’s approach in this case is one that both the Board and this Court have used in the past. For example, in Lizama v. Holder, 629 F.3d 440 (4th Cir.2011), the Court upheld the Board’s conclusion that Lizama’s proposed social group of “young, Americanized, well-off Salvadoran male deportees with criminal histories who oppose gangs” lacked particularity. Id. at 442. In so holding, the Court observed that the component parts of this group were “all amorphous characteristics that neither ‘provide an adequate benchmark for determining group membership,’ nor embody concrete traits that would readily identify a person as possessing those characteristics.” Id. at 447 (internal citation omitted). Indeed, Lizama specifically argued that the Board had erred by “dissectfing] [Liza-ma’s] social group into subparts, and fail[ing] to consider the group in its totality.” Id. Far from recognizing that approach as legal error, the Court rejected the argument for the simple reason that Lizama failed to “explain how viewing the above-mentioned terms in conjunction with each other makes the group any more particular[.] The truth remains that, as a whole, the group described is not narrow or enduring enough to clearly delineate its membership or readily identify its members.” Id. at 447-48. If the approach taken by the Board had been an error of law, Lizama would have recognized it as such and would not have applied the same approach in considering whether Lizama’s proposed group lacked particularity.
As another example, in Zelaya v. Holder, 668 F.3d 159 (4th Cir.2012), Zelaya asserted that he was eligible for asylum *901and withholding of removal based on his membership in the proposed group of “young Honduran males who (1) refuse to join the Mara Salvatrucha 13 gang (MS-13), (2) have notified the authorities of MS-13’s harassment tactics, and (3) have an identifiable tormentor within MS-13.” Id. at 162. In upholding the Board’s denial of relief based on the conclusion that Zelaya’s group was not cognizable under the INA, the Court held:
The critical problem with Zelaya’s proposed social group for purposes of seeking asylum is that it fails the BIA’s particularity requirement. First, as we have previously recognized, opposition to gangs is an amorphous characteristic providing neither an adequate benchmark for determining group membership nor embodying a concrete trait that would readily identify a person as possessing such a characteristic. Resisting gang recruitment is similarly amorphous, and the fact that Zelaya’s conduct in resisting recruitment included complaining twice to the police adds little to the particularity equation in the face of the common sense proposition that MS-13 would look unfavorably upon anyone who complained about its harassment tactics to the police. Similarly, the concept that a person who is victimized by one gang member more than by other gang members somehow serves to particularize all such persons into a targeted social group is just nonsensical.
Id. at 166-67. The Court thus considered the group as a whole, but in so doing addressed why its specific components lacked particularity as part of that whole. Similarly, here, the Board considered Temu’s proposed group and concluded that it consisted of criteria that would not make that group “particular.” (Cf. A.R. 4.) Far from being an error of law, that is precisely the inquiry the IJ and Board are charged with undertaking.
I conclude that the IJ and Board’s determinations after conducting that proper legal inquiry are not manifestly contrary to the law or an abuse of discretion. No adequate benchmark exists for determining whether an individual is a member of a group defined as “bipolar individuals in Tanzania who engage in erratic behavior.” Contrary to the majority’s claim, there is nothing inherent in this group’s description that limits a person’s erratic behavior to a diagnosis of bipolar disorder or vice versa. More to the point, bipolar disorder covers a wide spectrum of behaviors and tendencies, and “erratic behavior” is inherently subjective and amorphous. There is no discernible basis for readily identifying an individual as being part of the proposed group or not. See, e.g., Mendozar-Alvarez v. Holder, 714 F.3d 1161, 1164 (9th Cir.2013) (per curiam) (rejecting the proposed social groups of “all insulin-dependent diabetics or all insulin-dependent diabetics who suffer from mental illnesses” (whether the group contained the additional characteristics of an inability to work, lack of medical insurance, and a lack of money from other sources from which to pay for essential medication or not) as lacking particularity because the proposed group “include[s] large numbers of people with different conditions and in different circumstances ... [and] [individuals may have these conditions separately or in combination, and in varying degrees of severity” and thus are “far from a particular, discrete social group”); Matter of S-E-G-, 24 I. & N. Dec. at 585 (concluding that “male children who lack stable families and meaningful adult protection, who are from middle and low income classes, who live in territories controlled by the MS-13 gang, and who refuse recruitment” lacks particularity because “these characteristics remain amorphous *902[given that] ‘people’s ideas of what those terms mean can var/ ”).
The majority is correct that the IJ and Board must consider the scope of the group the petitioner actually proffers. See Maj. Op. at 895-96. And it is conceptually true that the INA does not require “that each individual trait must meet all the criteria for a ‘particular social group’ ” in order for the proposed group considered as a whole to satisfy the particularity requirement. See Maj. Op. at 896. But where none of the proposed group’s parts satisfy the requirement, then it will most frequently be the case that the whole cannot satisfy it either.
The majority’s reliance on Crespin-Valladares v. Holder, 632 F.3d 117 (4th Cir.2011), is, I believe, misplaced because there the petitioners’ proposed group consisted in part of family members, a trait that had previously been recognized as being a discrete group that was well-defined. Id. at 125. The Board’s error in that case was ignoring that component of the proposed group in conducting its original analysis and then re-characterizing the group on remand beyond its proposed scope. Id.; see also Zelaya, 668 F.3d at 166 (noting Crespin-Valladares’s holding that “the self-limiting nature of the family unit satisfied the particularity requirement”). Neither the IJ nor the Board engaged in similar conduct here; instead, they considered Temu’s proposed group according to each component he urged and found that this proposed group lacked the requisite particularity. For the reasons described above, I would hold that their analysis was not manifestly contrary to the law or an abuse of discretion.
III.
Because I would uphold the Board’s decision with respect to particularity, it follows that Temu’s proposed group would fail to satisfy all the required characteristics of a particular social group, and his claim would fail as a matter of law. Cf. Zelaya, 668 F.3d at 167. As such, I do not find it necessary to consider Temu’s remaining arguments. I would deny Temu’s petition for review, thereby affirming the Board’s decision to deny Temu’s application for asylum and withholding of removal under the INA. I therefore respectfully dissent.
. Citations to the adopted administrative record filed with this Court are denoted by "A.R.”
. The alternate diagnosis of schizophrenia is no longer part of Temu’s proposed social group.
.Chevron v. Nat’l Res. Defense Council, 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984).
. On appeal, Temu challenges the Board’s use of this three-part test in addition to its application of it, asserting that requiring "social visibility” is a marked departure from the Board's long-standing approach to defining a "particular social group” and that this Court should not afford this criteria Chevron deference. The Attorney General responds that Temu has waived appellate review of the Board’s use of the social visibility criteria because he failed to raise this issue in his appeal to the Board and thus has not exhaust his administrative remedies, as required by the INA.
The majority opinion does not address this issue and proceeds instead directly to the Board’s application of the social visibility cri-teña. I would expressly hold that the Court lacks jurisdiction to consider this aspect of Temu's claim. The INA expressly requires exhaustion of administrative remedies, 8 U.S.C. § 1252(d)(1), and as such there is no futility exception to this requirement and it must be strictly enforced. See Booth v. Churner, 532 U.S. 731, 741 n. 6, 121 S.Ct. 1819, 149 L.Ed.2d 958 (2001). And although this Court alone can resolve the question of Chevron deference, the Board could have considered anew whether its use of the three-part criteria was an appropriate interpretation of the INA and cured its purported lack of an explanation for why it was appropriate to use that construct had Temu raised that issue for the Board’s consideration.