PRECEDENTIAL
       UNITED STATES COURT OF APPEALS
            FOR THE THIRD CIRCUIT
                 _____________

                    No. 17-2031
                   _____________

            S.E.R.L.; Y.N.S.R.; Y.Y.R.L.,

                           Petitioners

                          v.

ATTORNEY GENERAL UNITED STATES OF AMERICA
              ______________

       On Petition for Review of an Order of the
         United States Department of Justice
           Board of Immigration Appeals

  (BIA-1:A206-800-692, A206-800-690, A206-800-691)
         Immigration Judge: Hon. John B. Carle
                  _______________

                    ARGUED
                   March 12, 2018

Before: JORDAN, KRAUSE, and GREENBERG,
                   Circuit Judges

                 (Filed: July 3, 2018)
                 _______________
David C. Bennion
1706 N. 2nd Street - #L4
Philadelphia, PA 19122

Elizabeth A. Cuneo
Russell H. Falconer [ARGUED]
Chelsea G. Glover
Gibson Dunn & Crutcher
2100 McKinney Avenue - #1100
Dallas, TX 75201

Charles Roth
Lisa Koop
Ashley Huebner
National Immigrant Justice Center
208 South LaSalle Street
Chicago, IL 60604
      Counsel for Petitioners

Jefferson B. Sessions, III
Chad A. Readler
Anthony P. Nicastro
Sheri R. Glaser [ARGUED]
United States Department of Justice
Office of Immigration Litigation
P. O. Box 878
Ben Franklin Station
Washington, DC 20044
       Counsel for Respondents




                             2
Blaine M. Bookey
Anne Dutton
Eunice Lee
Karen Musalo
Hastings College of the Law
Center for Gender & Refugee Studies
200 McAllister Street
San Francisco, CA 94102
      Amicus Counsel for American Immigration
      Lawyers Association, Center for Gender &
      Refugee Studies, and Hebrew Immigrant
      Aid Society PA

Kirsten L. Nathanson
Crowell & Moring
1001 Pennsylvania Avenue, NW
Washington, DC 20004

Emily T. Kuwahara
Daniel P. Wierzba
Crowell & Moring
515 S. Flower Street 40th Fl.
Los Angeles, CA 90071

Tu-Quyen Pham
Crowell & Moring
3 Park Plaza, 20th Fl.
Irvine, CA 92694
      Amicus Counsel for NIWAP Inc. and
      Pennsylvania Coalition Against Domestic Violence




                                3
                      _______________

                OPINION OF THE COURT
                    _______________


JORDAN, Circuit Judge

        In this immigration case, we consider the term
“particular social group,” which is part of the definition of
“refugee” in the Immigration and Nationality Act (“INA”), 8
U.S.C. § 1101(a)(42). We must decide whether a revised
interpretation of that term by the Board of Immigration
Appeals (the “BIA” or the “Board”) is reasonable and
therefore entitled to deference under the strictures of
Chevron, U.S.A., Inc. v. Natural Resources Defense Council,
Inc., 

(1984). Like other circuit courts, we had
dutifully deferred to the initial interpretation of that term
given by the Board in Matter of Acosta, 19 I. & N. Dec. 211,
233 (BIA 1985), overruled on other grounds by Matter of
Mogharrabi, 19 I. & N. Dec. 439 (BIA 1987). Fatin v. I.N.S.,

, 1239-40 (3d Cir. 1993). But, over time, the
Board began adding new requirements to its test for
determining whether an applicant had established the
existence of a particular social group and could thereby claim
refugee status. In Valdiviezo-Galdamez v. Attorney General,

(3d Cir. 2011), we concluded that the BIA had
departed from Acosta without a principled explanation and
that its new requirements for proving a particular social group
were incapable of consistent application. We therefore held
that its interpretation of “particular social group” was not
entitled to Chevron deference. 

                              4
       The BIA has since responded to our concerns. In a
pair of precedential decisions, Matter of M-E-V-G-, 26 I. &
N. Dec. 227 (BIA 2014), and Matter of W-G-R-, 26 I. & N.
Dec. 208 (BIA 2014), affirmed in part, vacated and
remanded in part on other grounds sub nom. Reyes v. Lynch,

(9th Cir. 2016), it articulated a three-part test
for proving the existence of a cognizable particular social
group. The test requires applicants to “establish that the
group [at issue] is (1) composed of members who share a
common immutable characteristic, (2) defined with
particularity, and (3) socially distinct within the society in
question.” M-E-V-G-, 26 I. & N. Dec. at 237. We now hold
that that statutory interpretation is entitled to Chevron
deference, and, applying the newly framed test here, we
conclude that substantial evidence supports the BIA’s
determination that the petitioner has not met its requirements.
Accordingly, we will deny the petition for review.

I.   BACKGROUND1

       S.E.R.L., a native of Honduras, seeks review of the
denial of her application for asylum and statutory withholding

       1
          We accept the agency’s factual findings as
conclusive, unless “any reasonable adjudicator would be
compelled to conclude to the contrary[.]”             8 U.S.C.
§ 1252(b)(4)(B). We therefore recite the facts as found by the
immigration judge, who determined that S.E.R.L. testified
credibly and afforded her testimony full evidentiary weight,
and as accepted by the BIA. We supplement the facts with
additional details found in S.E.R.L.’s affidavit and testimony,
where consistent with those findings.




                              5
of removal based on membership in a proposed particular
social group that she characterizes as “immediate family
members of Honduran women unable to leave a domestic
relationship[.]”2 (Opening Br. at 21.) She fears persecution
by two men, Jose Angel and Juan Orellana. Jose Angel
abducted, raped, and continues to stalk one of S.E.R.L.’s
daughters, K.Y.R.L. That daughter has already been granted
asylum in the United States. Juan Orellana is S.E.R.L.’s
stepfather and has repeatedly abused S.E.R.L.’s mother.
S.E.R.L. fears that if she is removed to Honduras, both men
will persecute her, Jose Angel because of her relationship to
her daughter, and Juan Orellana because of her relationship to
her mother.

      S.E.R.L. and two of her children fled here from
Honduras in 2014. Within a month of their unlawful arrival,
the Department of Homeland Security initiated removal
proceedings pursuant to INA § 212(a)(6)(A)(i).3 S.E.R.L.

      2
          This petition for review was filed on behalf of
S.E.R.L., as well as two of her minor children, Y.N.S.R. and
Y.Y.R.L.       The children are derivative applicants on
S.E.R.L.’s application for asylum and related relief, so we
will refer to S.E.R.L. as the petitioner, in the singular.
      3
          That statutory subsection provides:

      An alien present in the United States without
      being admitted or paroled, or who arrives in the
      United States at any time or place other than as
      designated by the Attorney General, is
      inadmissible.




                                6
conceded removability, and timely applied for asylum and
statutory withholding of removal.4 In support of her claims
for relief, she alleged past persecution and a fear of future
persecution based on the relationships just noted.

       An immigration judge (“IJ”) reviewed S.E.R.L.’s
application and conducted a merits hearing. Although finding
her credible, the IJ concluded that S.E.R.L. had not met her
burden to establish eligibility for any of the relief she had
requested. According to the IJ, S.E.R.L. had not established
past persecution or an objectively reasonable fear of future
persecution by Jose Angel, given that he had targeted
S.E.R.L.’s daughter, not her. Though crediting S.E.R.L.’s
testimony about Juan Orellana’s abuse of her mother and past
threats directed at S.E.R.L., herself, the IJ also noted that
S.E.R.L. said “her stepfather never physically harmed her.”
(Administrative Record (“AR”) at 86.) The IJ did not state
whether S.E.R.L. had established past persecution by Juan
Orellana.

       The IJ did say that, even if she had suffered past
persecution, S.E.R.L. failed to establish that the harm she
suffered was on account of a protected ground. Applying the
BIA’s newly clarified three-part test from M-E-V-G-, the IJ
rejected S.E.R.L.’s argument that “immediate family
members of Honduran women unable to leave a domestic
relationship” constituted a cognizable group. (AR at 89-90.)
According to the IJ, the group “lack[ed] the requisite level of


8 U.S.C. § 1182(a)(6)(A)(i).
      4
        S.E.R.L. also initially sought protection under the
Convention Against Torture but no longer pursues that relief.




                               7
particularity and social distinction” and thus failed the test’s
second and third requirements. (AR at 90.) The IJ also noted
that “asylum and refugee laws do not protect people from
general conditions of strife, such as crime and other societal
afflictions.” (AR at 90 (quoting M-E-V-G-, 26 I. & N. Dec. at
235).) Consequently, the IJ denied relief and ordered that
S.E.R.L. be removed.5

        She appealed that decision to the BIA. It too
concluded that she had not met her burden to establish
eligibility for either asylum or withholding of removal. It
agreed with the IJ’s conclusion that she had not established
past persecution by Jose Angel, and it further concluded that
she had not established past persecution by Juan Orellana,
because any threats he made “d[id] not rise to the level of
persecution[.]” (AR at 4.)

       The BIA also agreed that S.E.R.L.’s proposed
particular social group – immediate family members of
Honduran women unable to leave a domestic relationship –
lacked the requisite particularity and social distinction. As to

       5
         The IJ rejected three other particular social groups
proposed by S.E.R.L. as alternatives, including (1) Honduran
women unable to leave a domestic relationship, (2) immediate
family members of young Honduran women without a father
in the home, and (3) Honduran women who report gender-
based crimes to the police. S.E.R.L. has not challenged those
rulings in her petition to us and so those proposed groups are
not before us. See Frias-Camilo v. Att’y Gen., 

,
701 n.1 (3d Cir. 2016) (explaining that a petitioner’s failure to
challenge certain portions of the BIA’s decision results in
waiver).




                               8
particularity, the BIA observed that “[the] proposed group
could include individuals of any age, sex, or background, and
it is not limited to those who … take overt action to assist, or
are meaningfully involved with, the family member who is
unable to leave a domestic relationship.” (AR at 5.) The BIA
further “agree[d] that [S.E.R.L.] ha[d] not presented evidence
that this group is socially distinct within Honduran society, as
the record does not reflect that members of such a group
would be perceived, considered, or recognized in Honduras as
a distinct group[.]” (AR at 5.) Even assuming a cognizable
particular social group, the Board “discern[ed] no legal error
or clear factual error” in the IJ’s determination that S.E.R.L.
had not established a well-founded fear of future persecution
by Jose Angel. (AR at 5-6.) The Board did not, however,
reach the issue of future persecution by Juan Orellana.

       It turned last to the question of withholding of removal
and concluded that, “[i]nasmuch as [S.E.R.L.] has failed to
satisfy the lower burden of proof required for asylum, it
follows that she has failed to satisfy the more stringent
standard required for withholding of removal[.]” (AR at 6.)
The Board thus dismissed the appeal. S.E.R.L. has timely
petitioned for review.




                               9
II.   DISCUSSION6

       S.E.R.L. contends that she is entitled to asylum and
withholding of removal because she has established a well-
founded fear of future persecution on account of her
membership in a legally cognizable particular social group,
that again being “immediate family members of Honduran
women unable to leave a domestic relationship[.]” (Opening
Br. at 21.) The parties’ primary dispute is whether the BIA’s
revised interpretation of “particular social group,” as set forth
in Matter of M-E-V-G-, warrants Chevron deference.
S.E.R.L., supported by amici,7 asks us to reject the test from
M-E-V-G- because it is “deeply flawed,” “has no basis in the
asylum statute,” and fails to resolve the concerns raised in our

       6
         The BIA had jurisdiction under 8 U.S.C. § 1103 and
8 C.F.R. § 1003.1(b). Because “denial of … [an] applicant’s
petition for asylum, withholding of removal, and relief under
the CAT constitutes a final order of removal,” Shehu v. Att’y
Gen., 

, 656 (3d Cir. 2007) (internal quotation
marks and citation omitted), we have appellate jurisdiction
pursuant to 8 U.S.C. § 1252(a).
       7
          We thank the National Immigrant Women’s
Advocacy Project and the Pennsylvania Coalition Against
Domestic Violence, and the Center for Gender & Refugee
Studies, American Immigration Lawyers Association, and
Hebrew Immigrant Aid Society Pennsylvania for filing
amicus briefs in this matter, which have assisted our
consideration of the legal issues before us and also shine a
light on an issue of international concern: violence against
women, including Honduran women who intervene on behalf
of victims suffering from domestic abuse.




                               10
decision in Valdiviezo-Galdamez. (Opening Br. at 1-2.)
Instead, S.E.R.L. argues, we should continue to apply the test
from Matter of Acosta, which she claims to “satisf[y] … with
ease.” (Opening Br. at 22.) She also says that, in the event
the Board’s new interpretation is given deference, she has met
its particularity and social distinction requirements. Finally,
she contends that remand is required, if for no other reason,
because neither the IJ nor the BIA addressed whether she has
a well-founded fear of future persecution by Juan Orellana.

       Before we address those arguments, we first discuss
the governing legal principles and provide a review of our
Valdiviezo-Galdamez decision and the BIA’s response in
M-E-V-G-.

       A.     General Legal Principles

              1.     Standard of Review

       Whether a petitioner’s “proffered particular social
group is cognizable under [8 U.S.C. § 1101(a)(42)(A)] is a
question of law … subject to de novo review,” which, we
have said, is “subject to established principles of [Chevron]
deference[.]” Gomez-Zuluaga v. Att’y Gen., 

,
339 (3d Cir. 2008) (internal quotation marks and citation
omitted).8 More precisely, the existence of a cognizable

       8
         In cases like this one, we have often described the
governing standard of review as being “de novo, subject to
principles of Chevron deference.” See Mondragon-Gonzalez
v. Att’y Gen., 

, 158 (3d Cir. 2018) (“We accord
de novo review to questions of law, including the BIA’s
interpretation of the INA, subject to the deference dictated by




                              11
particular social group presents a mixed question of law and
fact, since the ultimate legal question of cognizability
depends on underlying factual questions concerning the group
and the society of which it is a part. Cf. 

(noting the “sparse” evidence supporting the
petitioner’s proposed particular social group, and concluding
that, even if cognizable, “the administrative record does not
establish that she is a member of [her proposed] group”). We
thus review de novo the ultimate legal conclusion as to the
existence of a particular social group, while we review the
underlying factual findings for “substantial evidence[.]” See
Lukwago v. Ashcroft, 

, 167 (3d Cir. 2003)
(reviewing the BIA’s statutory interpretation of “particular
social group” in accordance with Chevron principles, and
stating, “[o]n the other hand, we must treat the BIA’s findings
of fact as ‘conclusive unless any reasonable adjudicator
would be compelled to conclude to the contrary’” (quoting 8
U.S.C. § 1252(b)(4)(B))).

       Whether a petitioner has established membership in a
particular social group also involves agency fact-finding. 

178-79. “Our review is confined solely to the
administrative record,” 

, and
administrative findings of fact are “conclusive unless any
reasonable adjudicator would be compelled to conclude to the
contrary,” 

U.S.C. § 1252(b)(4)(B)). That
means that factual “determinations will be upheld if they are


Chevron[.]”). That may sound like a contradiction in terms.
What we mean is that we are required by Chevron principles
to defer to the BIA’s interpretation of the INA, when
reasonable, but we review de novo any legal challenge to the
application of that interpretation.




                              12
supported by reasonable, substantial, and probative evidence
in the record considered as a whole.” Kang v. Att’y Gen., 

, 164 (3d Cir. 2010) (citation omitted).

       Because here “the BIA adopted and affirmed the IJ’s
decisions and orders as well as [conducting] an independent
analysis, we review both the IJ’s and the BIA’s decisions and
orders.” Ordonez-Tevalan v. Att’y Gen., 

, 340-
41 (3d Cir. 2016). But we look to the IJ’s opinion “only
where the BIA has substantially relied on that opinion.”
Camara v. Att’y Gen., 

, 201 (3d Cir. 2009), as
amended (Nov. 4, 2009).

             2.     Asylum and Withholding of Removal

        Under the INA, the Attorney General has the
discretion to grant asylum to a removable alien, 8 U.S.C.
§ 1158(b)(1)(A), as long as the alien meets the INA’s
definition of “refugee.” That definition is as follows:

      Any person who is outside any country of such
      person’s nationality or, in the case of a person
      having no nationality, is outside any country in
      which such person last habitually resided, 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[.]

                             13
       A petitioner has the burden to establish that she is a
refugee, and thus eligible for asylum relief under the INA. 

One way of doing so is to show “a well-
founded fear of persecution on account of … membership in a
particular social group[.]” 

That is the
route S.E.R.L. chose to pursue, and so she bore the burden of
establishing the following elements: (1) a particular social
group that is legally cognizable; (2) membership in that
group; (3) a well-founded fear of persecution, which must be
subjectively genuine and objectively reasonable; and (4) a
nexus, or causal link, between the persecution and
membership in the particular social group. 

.9

      As for withholding of removal, under 8 U.S.C.
§ 1231(b)(3), an alien must “establish a ‘clear probability of
persecution,’ i.e., that it is more likely than not, that s/he
would suffer persecution upon returning home.” Valdiviezo-

       9
         Our decision in Fatin v. I.N.S. lays these out as three
elements, combining into one what we have noted here as
elements (3) and (4). See 

(“The alien
must … (3) show that he or she would be persecuted or has a
well-founded fear of persecution based on that
membership.”). Because the issue of nexus between alleged
persecution and membership in a particular social group is
sometimes the focus of dispute, see 

(reviewing the BIA’s conclusion that the petitioner
“failed … [to] show her political opinion or her particular
social group constituted ‘at least one central reason’ for her
persecution”), we have thought it best to frame it here as a
separate element of proof.




                              14


(citing I.N.S. v. Stevic, 

, 429-30 (1984)).      “Since [that] standard is more
demanding than that governing eligibility for asylum, an alien
who fails to qualify for asylum is necessarily ineligible for
withholding of removal.” 

Decision in Valdiviezo-Galdamez and the
              BIA’s Response in Matter of M-E-V-G-

       In Valdiviezo-Galdamez, we reviewed at length the
BIA’s evolving efforts to interpret the term “particular social
group,” beginning with the definition it set forth in Matter of
Acosta. We need not fully repeat that history here but, for
purposes of our analysis, will summarize a few important
points from the pertinent decisions of the BIA.

        From 1985 to 2006, the Board interpreted “particular
social group” to mean “a group of persons all of whom share
a common, immutable characteristic.” 

(quoting Acosta, 19 I. & N. Dec. at 233).
That standard became known as the Acosta test. It was rooted
in the interpretive doctrine of ejusdem generis, which teaches
that words in a list should be understood as referring to things
of the same general class or kind. Acosta, 19 I. & N. Dec. at
233. In the context of the statutory definition of “refugee,”
that means that the term “particular social group” should be
understood as being akin to the other characteristics listed in
the definition, namely race, religion, nationality, and political
opinion.        8 U.S.C. § 1101(a)(42)(A); see also

1231(b)(3). According to the BIA, all of
those focus on “an immutable characteristic,” which the BIA
explained includes both those characteristics that are
technically “immutable” as well as those a person “should not




                               15
be required” to change “as a matter of conscience” to avoid
persecution. Acosta, 19 I. & N. Dec. at 233-34. Thus, the
Acosta test required members of a “particular social group” to
have “a common, immutable characteristic” that “the group
either cannot change, or should not be required to change
because it is fundamental to their individual identities or
consciences.” 

listed examples of innate
characteristics, like “sex, color, or kinship ties[.]” 

noted that, in certain circumstances, “a shared past experience
such as former military leadership or land ownership” could
be the defining characteristic of a cognizable “particular
social group,” but such determinations would be made “on a
case-by-case basis.”10 

employing the Acosta test, the BIA
recognized several particular social groups based on
discernable and immutable characteristics. For example, in In
re H-, 21 I. & N. Dec. 337 (BIA 1996), it accepted “kinship”
as an immutable characteristic, concluding that “[t]he record
before us makes clear not only that the Marehan [– a familial
sub-clan in Somalia –] share ties of kinship, but that they are
identifiable as a group based upon linguistic commonalities.”

Importantly, however, in other cases, the Board
accepted particular social groups that did not share such
plainly discernable characteristics. In Matter of Fuentes, 19 I.

       10
           In Acosta, itself, the petitioner had claimed
persecution on account of his membership in a group of San
Salvador taxi drivers who refused to participate in guerrilla-
sponsored work stoppages. 

The BIA rejected
that proposed group under the “immutable characteristic” test,
concluding that a taxi driver could change his occupation and
avoid the danger he faced. 

                               16
& N. Dec. 658 (BIA 1988), it recognized the status of being a
former policeman as an innate characteristic and, although not
definitively reaching the issue in that case, it stated that
mistreatment because of that status could constitute
persecution on account of political opinion or membership in
a particular social group. 

In Matter of Toboso-
Alfonso, 20 I. & N. Dec. 819 (BIA 1990), it accepted a
particular social group of homosexuals in Cuba. 

23. And in In re Kasinga, 21 I. & N. Dec. 357 (BIA 1996), it
accepted the particular social group of “young women of the
Tchamba-Kunsuntu Tribe who have not had [female genital
mutilation], as practiced by that tribe, and who oppose the
practice.” 

       Eventually, the BIA determined that the Acosta test
had proven to be over-inclusive and unworkable, in part
because it encompassed virtually any past acts or experiences,
since the past cannot be changed and is, by definition,
immutable. Thus, in 1999, the BIA began supplementing the
Acosta test with additional requirements.          Valdiviezo-

. For example, in In re R-A-,
22 I. & N. Dec. 906 (BIA 1999; A.G. 2001), remanded for
reconsideration in Matter of R-A-, 24 I. & N. Dec. 629 (A.G.
2008),11 it took issue with particular social groups that were
“defined principally, if not exclusively, for the purposes of

      11
         In that case, the Attorney General remanded the case
for reconsideration in light of the Board’s intervening
decisions. See Matter of R-A-, 24 I. & N. Dec. at 630 (citing
Matter of E-A-G-, 24 I. & N. Dec. 591 (BIA 2008); Matter of
S-E-G-, 24 I. & N. Dec. 579 (BIA 2008); In re A-M-E- &
J-G-U-, 24 I. & N. Dec. 69 (BIA 2007); In re C-A-, 23 I. &
N. Dec. 951 (BIA 2006)).




                             17
[litigation] … without regard to the question of whether
anyone in [a given country] perceives [those] group[s] to exist
in any form whatsoever.” 

Although the Board
maintained that the Acosta test was the starting point for
assessing particular social groups, it said that the test would
no longer be the ending point. 

Other factors
would be considered, including whether the alleged defining
characteristic of the social group is important within the
society in question and whether that society understands or
recognizes the proposed social group as a distinct segment of
the population. 

        By 2006, the BIA appeared to have transformed its
requirements for establishing a particular social group into a
new three-part test: (1) the original Acosta test, requiring
members to have a common, immutable characteristic; (2)
social visibility, meaning that members of the social group
are visible and recognizable by others in the society in
question; and (3) particularity, meaning that the group has
defined boundaries. See 

(citing a pair of BIA cases as establishing the social
visibility and particularity requirements, In re C-A-, 23 I. &
N. Dec. 951 (BIA 2006), aff’d sub nom. Castillo-Arias v.
Att’y Gen., 

(11th Cir. 2006), and In re A-M-E-
& J-G-U-, 24 I. & N. Dec. 69 (BIA 2007), aff’d sub nom.
Ucelo-Gomez v. Mukasey, 

(2d Cir. 2007)).
Although several of our sister courts of appeals gave Chevron
deference to that interpretation,12 we, along with the Seventh

      12
          See 

n.16
(noting that the First, Second, Eighth, Ninth, and Eleventh
Circuits had accepted the BIA’s three-part definition without
issue).




                              18
Circuit, rejected the BIA’s social visibility and particularity
requirements. 

see also Benitez Ramos v.
Holder, 

(7th Cir. 2009); Gatimi v. Holder, 

(7th Cir. 2009).

        In Valdiviezo-Galdamez, we took issue with the BIA’s
departure from 

; see also 

(Hardiman, J., concurring). Two specific concerns
animated our analysis. First, we said that the BIA had applied
the “social visibility” requirement in an “inconsistent”
manner. 

Specifically, we expressed concern
that, in cases like In re C-A-, the Board had referred to “social
visibility” as “recognizability” and as “involv[ing]
characteristics that were highly visible and recognizable by
others in the country in question,” 

(quoting
C-A-, 23 I. & N. Dec. at 959-60), yet in other cases, it had
accepted particular social groups for refugee status based on
internal characteristics that lacked any apparent visibility,
absent self-disclosure, including “women who are opposed to
female genital mutilation (Matter of Kasinga), homosexuals
required to register in Cuba, (Matter of Toboso-Alfonso), and
former members of the El Salvador national police (Matter of
Fuentes).” 

        We cited the Seventh Circuit’s criticism that “[o]ften it
is unclear whether the Board is using the term ‘social
visibility’ in the literal sense, or in the ‘external criterion’
sense, or even whether it understands the difference.” 

(quoting Benitez 

). Because the
BIA had applied the social visibility requirement
inconsistently, we concluded that it was “an unreasonable
addition to the requirements for establishing refugee status




                               19
where that status turns upon persecution on account of
membership in a particular social group.” 

        Second, we said that social visibility and particularity
“appear to be different articulations of the same concept,” 

at least as the BIA had defined them in prior
decisions. 

To illustrate the point, we quoted the
decision in Matter of S-E-G-, 24 I. & N. Dec. 579 (BIA
2008), in which the BIA described the “essence” of
particularity as an assessment of “whether the proposed group
can accurately be described in a manner sufficiently distinct
that the group would be recognized, in the society in question,
as a discrete class of persons,” and noted that the size of the
group “may be an important factor[.]” 

(quoting
S-E-G-, 24 I. & N. Dec. at 584). The BIA did go on to say
that “the key question is whether the proposed description is
sufficiently particular, or is too amorphous ... to create a
benchmark for determining group membership,” 

original) (quoting S-E-G-, 24 I. & N. Dec. at
584), but what it said about the essence of particularity led us
to reject the requirement as both confusing and “little more
than a reworked definition of ‘social visibility[.]’” 

        Having decided that “the BIA’s requirements that a
‘particular social group’ possess the elements of ‘social
visibility’ and ‘particularity’ [were] inconsistent with prior
BIA decisions” and repetitive, we then held that they were not
entitled to Chevron deference. 

expressly noted
that the BIA was free to depart from or change its
interpretation of “particular social group,” and that a new
view could be entitled to deference if supported by a
“principled reason” and explanation for any new




                               20
requirements.13 

aftermath of Valdiviezo-Galdamez,
we continued to apply the BIA’s original Acosta test. See,
e.g., Garcia v. Att’y Gen., 

, 504 n.5 (3d Cir.
2011) (“Until the BIA provides an analysis that adequately
supports its departure from Acosta, we remain bound by the
well-established definition of ‘particular social group’ found
in Fatin [where we adopted the Acosta test].”).

       The BIA promptly responded to our concerns and
announced a revised interpretation of “particular social
group” in Matter of M-E-V-G-, which it also applied in a
companion case pending in the Ninth Circuit, Matter of
W-G-R-.      The Board adhered to its more restrictive
interpretation of particular social group, and it clarified the
three requirements that an applicant for asylum or
withholding of removal must satisfy to establish a cognizable

      13
           In a concurring opinion, Judge Hardiman said:

      [T]he only problem that I find with the BIA’s
      evolving approach to ‘particular social group’
      cases is that the Board has failed to
      acknowledge a change in course and
      forthrightly address how that change affects the
      continued validity of conflicting precedent.
      Accordingly, remand is necessary so the Board
      can either choose between its reasonable new
      requirements and its older but equally
      reasonable precedents, or reconcile the two
      interpretations in a coherent way.

(Hardiman, J.,
concurring).




                              21
particular social group. As stated in M-E-V-G-, an applicant
must “establish that the [proposed] group is (1) composed of
members who share a common immutable characteristic,
(2) defined with particularity, and (3) socially distinct within
the society in question.” 26 I. & N. Dec. at 237.

       The BIA reviewed its prior efforts to outline what
constitutes a “particular social group.” 

that the
addition of “particularity” and “social distinction” as required
elements is both “consistent with … the language of the
[INA]” as well as consistent with the interpretation “set forth
in Matter of Acosta[.]” Id.; see also 

(citing earlier
cases, and stating that it would continue to “adhere to the
social group requirements announced in” its prior decisions).
It explained that the INA’s “enumerated grounds of
persecution have more in common than simply describing
persecution aimed at an immutable characteristic. They have
an external perception component within a given society,
which … separates various factions within a particular
society.”    

     In the Board’s view, adding
“particularity” and “social distinction” as requirements for
proving a particular social group became necessary, based on
its experience in cases since Acosta. 

       The “particularity” requirement, it said, “is included in
the plain language of the [statute] and is consistent with the
specificity by which race, religion, nationality, and political
opinion are commonly defined.” 

“Particularity” is
largely definitional, ensuring that the characteristics defining
a group “provide a clear benchmark for determining who falls
within the group.” 

explained that particularity
requires the group to be “discrete and have definable
boundaries” that are not “amorphous, overbroad, diffuse, or




                              22
subjective,” ensuring that an immutable characteristic is
“sufficiently precise to define a particular social group.” 

distinction” requirement, it said, was a
reworking of the social visibility requirement and was
intended to resolve any “misconception” that literal visibility
was meant. 

According to the BIA, social visibility
“was never meant to be read literally.” 

The
change in terminology to “social distinction” was made to
“clarif[y] that social visibility does not mean ‘ocular’
visibility – either of the group as a whole or of individuals
within the group – any more than a person holding a protected
religious or political belief must be ‘ocularly’ visible to others
in society.” 

Board explained, “[t]o be
socially distinct, a group need not be seen by society; rather,
it must be perceived as a group by society.” 

noted our concern about the inconsistent
application of the former “social visibility” requirement, and
described why it viewed the revised social distinction
requirement as nevertheless being an appropriate approach. It
stated:

       It may not be easy or possible to identify who is
       opposed to [female genital mutilation], who is
       homosexual, or who is a former member of the
       national     police.         These     immutable
       characteristics are certainly not ocularly visible.
       Nonetheless, a society could still perceive
       [members of those groups] to comprise a
       particular social group for a host of reasons,
       such as sociopolitical or cultural conditions in
       the country.




                               23


        For that reason, it said, “the fact that members of a
particular social group may make efforts to hide their
membership in the group to avoid persecution does not
deprive the group of its protected status as a particular social
group.” 

also directly addressed its prior
decision in In re C-A-, stating, “to the extent that [the
decision] has been interpreted as requiring literal or ‘ocular’
visibility, we now clarify that it does not.” 

        The BIA then answered our concern that particularity
and social visibility, now recast as social distinction, are not
discernibly different. 

Although acknowledging
that “there is considerable overlap” between particularity and
social distinction, the BIA explained its view that they are
both different and necessary. 

that, although
relying on an overlapping body of evidence, “each
emphasize[s] a different aspect of a particular social group.”

“Particularity” addresses “the ‘outer limits’ of a
group’s boundaries and is definitional in nature,” whereas
“social distinction” focuses on “whether the people of a given
society would perceive a proposed group as sufficiently
separate or distinct[.]” 

BIA also took the opportunity to
emphasize that “a group’s recognition for asylum purposes is
determined by the perception of the society in question, rather
than by the perception of the persecutor.” 

There
must be a distinction, the Board explained, between the INA’s
requirement that an applicant “establish[] the existence of one
of the enumerated grounds,” including “particular social




                              24
group,” and the INA’s nexus requirement, which addresses
whether an applicant has suffered persecution “on account of”
that enumerated ground. 

to the extent
indicative of society’s views as a whole, the Board stated that
“persecutory conduct alone cannot define [a particular social]
group.” 

background in mind, we now turn to the
main dispute in this case – whether the revised test for
determining the cognizability of a particular social group
resolves the concerns we raised in Valdiviezo-Galdamez and
is therefore entitled to Chevron deference.

       C.     The BIA’s Revised Interpretation of
              “Particular Social Group” is Entitled to
              Chevron Deference

       “Congress has charged the Attorney General with
administering the INA,” who has chosen to delegate that
authority to the BIA. Negusie v. Holder, 

, 516-
17 (2009). And, “[c]onsistent with the rule in Chevron …,
the BIA is entitled to deference in interpreting ambiguous
provisions of the INA.” 

The Supreme Court has
instructed that “[j]udicial deference in the immigration
context is of special importance, for executive officials
exercise especially sensitive political functions that implicate
questions of foreign relations.” 

(internal quotation
marks and citation omitted); see also 

the
judiciary is not well-suited to assume primary responsibility
for assessing important diplomatic factors). Hence, “the BIA
should be accorded Chevron deference as it gives ambiguous
statutory terms ‘concrete meaning through a process of case-




                              25
by-case adjudication.’” 

v. Aguirre-
Aguirre, 

, 425 (1999)).

       The familiar Chevron two-step analysis thus applies
with full force in the immigration context.             When
“considering an interpretation adopted by the Board, we must
ask ‘whether Congress has directly spoken to the precise
question at issue.’” 

(quoting 

). “If it has not, we may not ‘simply impose
[our] own construction on the statute.’” 

original) (quoting 

). “Rather, if the
statute is silent or ambiguous with respect to the specific
issue, the question for the court is whether the agency’s
answer is based on a permissible construction of the statute.”
Id. (quoting 

).

        Our case law has already established that the term
“particular social group” is undefined in the statute, and its
meaning is unclear. We have observed that “[b]oth courts
and commentators have struggled to define ‘particular social
group.’ Read in its broadest literal sense, the phrase is almost
completely open-ended.” 

(quoting 

). The statutory language
is “not very instructive” and there is scant evidence of
legislative intent. Id. (quoting 

). Thus,
the question before us now, as in the past, is whether the
Board’s interpretation of that ambiguous term is a reasonable
one.14

       14
           Although initially contending that the BIA’s new
interpretation fails Chevron step one, S.E.R.L. acknowledged
at oral argument that the Chevron framework applies, and that
the term “particular social group” is ambiguous. We agree




                              26
       S.E.R.L. contends that the BIA’s change in
nomenclature from “social visibility” to “social distinction” is
the only change the BIA has made to its test for assessing a
“particular social group,” and, she says, that is a “distinction
without a difference.” (Reply Br. at 5.) According to
S.E.R.L., our decision in Valdiviezo-Galdamez forecloses
application of the “particularity” and “social distinction”
requirements.     She also argues that the BIA plainly
acknowledges that it has not changed course, nor has it
provided a “principled” explanation for why it continues to
impose criteria we rejected in Valdiviezo-Galdamez.
(Opening Br. at 31.)

       In addition, those who have filed amicus briefs in this
case point out that the BIA’s decisions in M-E-V-G- and
W-G-R- could be read as inconsistent with certain other BIA
decisions and contrary to the canon of ejusdem generis.
Amici note, for example, that in W-G-R-, the BIA concluded
that “‘former members of the Mara 18 gang in El Salvador
who have renounced their gang membership’ does not
constitute a particular social group” in part because “the
group could include persons of any age, sex, or background.”
26 I. & N. Dec. at 221. Yet, even though the groups varied
significantly across age, sex, and background, the BIA has
also held that “Filipinos of Chinese [a]ncestry” constituted a
“particular social group,” In re V-T-S-, 21 I. & N. Dec. 792,
798 (BIA 1997), and that “former member[s] of the national


and thus proceed directly to step two. See Valdiviezo-

(rejecting the petitioner’s
argument that the BIA’s particularity and social visibility
requirements were contrary to the intent of the INA).




                              27
police” in El Salvador, Fuentes, 19 I. & N. Dec. at 662,
likewise could be cognizable.15 And although the BIA
expressly justified its new requirements as “[c]onsistent with
the interpretive canon ‘ejusdem generis,’” M-E-V-G-, 26 I. &
N. Dec. at 234, amici highlight that some of the enumerated
grounds for persecution, including “political opinion,” and
“religion,” 8 U.S.C. § 1101(a)(42)(A), may themselves be
thought of as amorphous, diffuse, or subjective and therefore
as insufficient bases for PSGs under M-E-V-G-’s
requirements.

       Those critiques raise legitimate concerns. The BIA
has chosen to maintain a three-part test for determining the
existence of a particular social group, and it has discussed
how the revised particularity and social distinction
requirements are not a departure from but a ratification of
requirements articulated in its prior decisions. M-E-V-G-, 26
I. & N. Dec. at 234. And the arguable inconsistencies in its
precedent highlight the risk that those requirements could be
applied arbitrarily and interpreted to impose an unreasonably
high evidentiary burden, especially for pro se petitioners, at
the threshold. At the same time, however, we recognize that
M-E-V-G- is a relatively recent decision and clarity and
consistency can be expected to emerge with the accretion of
case law. That process is aided by M-E-V-G- itself, which

      15
         Although S.E.R.L. also relies heavily on Matter of
A-R-C-G-, 26 I. & N. Dec. 388 (BIA 2014), where the BIA
had held that “married women in Guatemala who are unable
to leave their relationship” constituted a particular social
group, the Attorney General recently issued a decision
overruling A-R-C-G-. See Matter of A-B-, 27 I. & N. Dec.
316 (A.G. 2018).




                             28
addressed the specific concerns we raised in Valdiviezo-
Galdamez, and explained why the particularity and social
distinction requirements are different from one another and
necessary. We now consider each of those requirements,
beginning with social distinction, to explain why,
notwithstanding our concerns, we conclude that the
requirements are reasonable and warrant Chevron deference.

             1.     Social Distinction

       “Social distinction” means social recognition, or
“whether the people of a given society would perceive a
proposed group as sufficiently separate or distinct[.]”
M-E-V-G-, 26 I. & N. Dec. at 241. The BIA has clarified that
“social distinction” is not a matter of being “seen” by society
in an “ocular” sense, as one might have understood from
decisions applying the old “social visibility” factor. 

The change in terminology from “social visibility” to
“social distinction” was intended to resolve any
“misconception” that literal visibility was a requirement. 

As defined in M-E-V-G-, social distinction accounts
for the particular social groups that the BIA has recognized in
the past and wishes to continue to recognize, including those
whose members share an immutable, though not literally
visible, characteristic.16 See 

(addressing

      16
          S.E.R.L. argues that the Board’s interpretation in
M-E-V-G- is unreasonable because the petitioners that
prevailed in several earlier cases could not have satisfied the
new test on the record before the agency in those cases. If
that were the litmus for assessing an agency’s revised
interpretation, however, then its first interpretation would be
all but set in stone. The Supreme Court has expressly




                              29
Kasinga, Toboso-Alfonso, and Fuentes). The Board thus
addressed our concern in Valdiviezo-Galdamez that it had
seemingly defined “social visibility” in “the literal sense” and
had been applying it inconsistently. 

(quoting Benitez 

).

       S.E.R.L. nevertheless suggests that by defining the
“social distinction” factor as based on the perception of the
society in question rather than by the perception of the
persecutor, the Board has impermissibly conflated the INA’s
“particular social group” and “nexus” requirements, rendering
the test set forth in Matter of M-E-V-G- an unreasonable
interpretation. We disagree, and we are not the first court to
do so.

       Reviewing the companion case to M-E-V-G-, the Ninth
Circuit considered and rejected a similar challenge. In Reyes
v. Lynch, the court concluded that “the ‘social distinction’
requirement is not redundant in light of the ‘nexus’
requirement for asylum and withholding claims.” 

, 1136 (9th Cir. 2016) (reviewing W-G-R-), cert. denied
sub nom. Reyes v. Sessions, 

(2018). “Rather
than conflate the ‘social distinction’ and ‘nexus’
requirements,” the court said, “the BIA’s reasoning reflects


rejected such a rigid standard and has acknowledged that an
agency “must consider varying interpretations and the
wisdom of its policy on a continuing basis,” Rust v. Sullivan,

, 186 (1991) (citation omitted), and, when it
concludes that deviation is required, that it “display
awareness that it is changing position” and “provide reasoned
explanation” for the change, FCC v. Fox Television Stations,
Inc., 

, 515 (2009) (emphasis omitted).




                              30
an appreciation of the need to distinguish between the
showing an applicant must make in order to demonstrate
membership in a ‘particular social group’ and the showing
that is necessary to demonstrate that he was persecuted, or
fears persecution, ‘on account of’ that membership.” Id.; see
also I.N.S. v. Elias-Zacarias, 

, 483 (1992)
(explaining that the INA “makes motive critical” and
requiring that an asylum applicant present evidence of his
persecutors’ motives to satisfy the “nexus” requirement).
That reasoning is entirely persuasive.

       It is well within the bounds of reasonableness for the
BIA to interpret the term “particular social group” in the INA
as requiring evidence that the society in question recognizes a
proposed group as distinct. The persecutor’s motive may be
relevant but is not alone sufficient in that regard. See
M-E-V-G-, 26 I. & N. Dec. at 243. Otherwise, every act of
persecution could be claimed as being on the basis of a
protected ground, since the internal motivations of a
persecutor are likely to be more obscure than are the
perceptions of a society generally. Also, one bad actor’s
twisted views should not be attributed to a whole society. We
therefore agree with the Ninth Circuit that the BIA’s
interpretation better maintains the distinction between
“particular social group” and the “nexus” requirement.

       Finally, although we are cognizant of arguable
inconsistencies in its application to date and the need for
careful review by the BIA and this Court to ensure a fair and
principled approach, we reject the suggestion by S.E.R.L. and
amici that the BIA’s social distinction requirement is
categorically incapable of rational application and that the
BIA has failed to “provide meaningful guidance about how




                              31
one would establish social distinction.” (Reply Br. at 17.) In
M-E-V-G-, the BIA described the kind of evidence that a
petitioner could rely on, stating “[e]vidence such as country
conditions reports, expert witness testimony, and press
accounts of discriminatory laws and policies, historical
animosities, and the like may establish that a group exists and
is perceived as ‘distinct’ … in a particular society.” 26 I. &
N. Dec. at 244. We do not read that list as exclusive, and it is
not unlike evidence the Board relies on in petitions alleging
persecution on account of other enumerated grounds. See
generally Sheriff v. Att’y Gen., 

, 592 (3d Cir.
2009) (discussing regulations “explicitly envision[ing] that
the BIA will consider Country Reports” and other official
documents); 

, 177 (noting the
use of testimonial, documentary, and expert evidence). Thus,
we conclude that the social distinction requirement is a
reasonable feature of the BIA’s interpretation of “particular
social group.”

              2.     Particularity

       Likewise, the particularity requirement is reasonable.
The word “particular” is in the text of the statute, 8 U.S.C.
§ 1101(a)(42)(A), and it is sensible to construe that word as
requiring an alleged social group to have “discrete and …
definable boundaries” that are not “amorphous, overbroad,
diffuse, or subjective,” M-E-V-G-, 26 I. & N. Dec. at 239, so
as to provide a clear standard for determining who is a
member of it, W-G-R-, 26 I. & N. Dec. at 214. The BIA has
explained that the “particularity requirement … clarif[ies] the
point, at least implicit in earlier case law, that not every
immutable characteristic is sufficiently precise to define a
particular social group.” 

For example, in Escobar




                              32
v. Gonzales, 

(3d Cir. 2005), a proposed
particular social group defined by “[p]overty, homelessness,
and youth” was held to be “too vague and all encompassing”
to set discernible parameters. 

       Given its explicit roots in the statute and the sensible
explanation of a need for some measure of definitional
precision, the particularity requirement is also a reasonable
feature of the BIA’s interpretation of “particular social
group.”


       17
          See also, e.g., S-E-G-, 24 I. & N. Dec. at 584-85
(stating that “the key question is whether the proposed
description is sufficiently particular, or is too amorphous ... to
create a benchmark for determining group membership,” and
rejecting the proposed group of “male children who lack
stable families and meaningful adult protection, who are from
middle and low income classes, who live in the territories
controlled by the MS-13 gang, and who refuse recruitment”
as too amorphous (alteration in original) (internal quotation
marks and citation omitted)); A-M-E- & J-G-U-, 24 I. & N.
Dec. at 76-77 (explaining that “affluent Guatemalans” did not
qualify as a particular social group in part because the
“characteristic of wealth or affluence is simply too subjective,
inchoate, and variable to provide the sole basis for
membership”); C-A-, 23 I. & N. Dec. at 959, 961 (rejecting a
proposed group of “noncriminal drug informants working
against the Cali drug cartel” due, in part, to the fact that the
distinction between government informants who had been
compensated for their services and those who acted out of
civic motives was not sufficient to carve out a particular
“subgroup” of uncompensated informants).




                               33
              3.     The BIA has adequately distinguished
                     social distinction and particularity

       We had expressed concern in Valdiviezo-Galdamez
that “social visibility” (now “social distinction”) and
“particularity” were really two ways of saying the same 

. But the BIA has adequately articulated why
it deems the ideas to be separate and why both are needed. In
the BIA’s reasoning, “social distinction” works to narrow the
universe of “particular social groups” to those whose
members are seen to be “distinct” or “other,” like the
distinctiveness inherent in the other enumerated grounds of
race, religion, nationality, and political opinion, while
“particularity” ensures that a group has “discrete …
boundaries” capable of a common, accepted definition.
M-E-V-G-, 26 I. & N. Dec. at 239-40, 244. We agree that
particularity and social distinction address different aspects of
whether an applicant has established a particular social
group.18 Although they may often involve similar evidence,
which the BIA readily acknowledges, 

that alone is
not a basis to reject them as being indistinguishable.

       Some overlap is to be expected, given that each
requirement is meant to illuminate whether a particular social
group exists in the society in question. See 

overlap because the overall definition is
applied in the fact-specific context of an applicant’s claim for

       18
          The Ninth Circuit has also concluded that the two
requirements are sufficiently distinct. See 

(discussing differences between social distinction
and particularity).




                               34
relief.”). But particularity and social distinction are different
in an important respect: the former is essentially an objective
inquiry, asking whether a reasonable person could look at the
proposed definition of a social group and determine who falls
within it, whereas the latter poses a more subjective question,
whether the alien’s home society actually does recognize that
group as being a “distinct” and identifiable group. Inquiring
separately about objective and subjective perspectives is a
familiar task in the law19 and is not out of bounds in this
context. For example, “[t]he well-found fear of persecution
standard involves both a subjectively genuine fear of
persecution and an objectively reasonable possibility of
persecution.” 

. In
short, we are satisfied that the BIA has explained why the two
requirements are not really just the same thing done over.

        The BIA has also explained why it views the addition
of “social distinction” and “particularity” as necessary
limitations on the Acosta test. It noted its concern that
Acosta’s immutable characteristic requirement resulted in
“confusion and a lack of consistency as adjudicators struggled
with various possible social groups, some of which appeared
to be created exclusively for asylum purposes.” M-E-V-G-,
26 I. & N. Dec. at 231. The additional requirements of social

       19
          See generally Kyllo v. United States, 

, 33
(2001) (describing Fourth Amendment inquiry as involving
“a subjective expectation of privacy that society recognizes as
reasonable”); United States v. Elonis, 

, 596 (3d
Cir. 2016) (construing 18 U.S.C. § 875(c)’s prohibition on
transmitting communications containing a threat to injure
another as including “both a subjective and objective
component”), cert. denied 

(2017).




                               35
distinction and particularity arose from the BIA’s experience
adjudicating prior cases and its desire to give further
guidance. When, in Valdiviezo-Galdamez, we remanded for
the Board to give a “principled reason” and explanation for
the added requirements, we indeed hoped to receive what we
asked for and did not intend to foreclose any additions to the
original Acosta test. 

see also 

(Hardiman,
J., concurring) (stating that “remand is necessary so the Board
can either choose between its reasonable new requirements
and its older but equally reasonable precedents, or reconcile
the two interpretations in a coherent way”); cf. 

(Scalia, J., and Alito, J., concurring) (“I would not
agree to remand if I did not think that the [BIA] has the
option of adhering to its decision. The majority appears to
leave that question undecided[.]”). S.E.R.L. is thus mistaken
in reading Valdiviezo-Galdamez as precluding the three-part
test the BIA adopted in M-E-V-G-.

       We are not alone in deferring to the BIA’s better
explained interpretation of “particular social group.” Since
we issued our decision in Valdiviezo-Galdamez, the majority
of our sister circuits have applied the test from M-E-V-G-,
including the First, Second, Fourth, Fifth, Sixth, Eighth,
Tenth, and Eleventh Circuits.20 Moreover, in Reyes, the

       20
           See Perez-Rabanales v. Sessions, 

, 66
(1st Cir. 2018) (applying BIA’s interpretation in M-E-V-G-
and rejecting proffered social group of “Guatemalan women
who try to escape systemic and severe violence but who are
unable to receive official protection”); Pacas-Renderos v.
Sessions, 691 F. App’x 796, 804 (4th Cir. 2017) (applying
criteria from M-E-V-G-); Hernandez-De La Cruz v. Lynch,

, 786-87 & n.1 (5th Cir. 2016) (endorsing BIA’s




                               36
Ninth Circuit expressly endorsed the BIA’s interpretation of
“particular social group” and granted it Chevron deference.

(concluding that “particularity” and
“social distinction” are reasonable requirements). The wide
acceptance of the BIA’s revised test from M-E-V-G-, and, in
particular, the Ninth Circuit’s analysis of the companion case,
W-G-R-, constitute persuasive support for our conclusion
today. Cf. In re Grossman’s Inc., 

, 121 (3d Cir.
2010) (en banc) (explaining that the “widely held views [of
other circuit courts] impel us to consider whether the
reasoning applied by our colleagues elsewhere is
persuasive”). Independent of Chevron, we are constrained to
acknowledge again that our role in the process of construing
the term “particular social group” is rightly limited. As the
Supreme Court has noted, courts are neither policy-makers
nor diplomats; we are ill-suited for those roles. 

. Immigration policy properly resides with the
elected branches of government.

       And, of course, we are not operating independently of
the rule in Chevron. 

. The Chevron
doctrine of deference to federal agencies is open to question,


interpretation); Zaldana Menijar v. Lynch, 

, 498-
99 (6th Cir. 2015) (same); Rodas-Orellana v. Holder, 

, 992 (10th Cir. 2015) (concluding that M-E-V-G-
and W-G-R- are “consistent with [the court’s] past
interpretation of social visibility”); Juarez Chilel v. Holder,

, 855 (8th Cir. 2015) (endorsing BIA’s
interpretation); Paloka v. Holder, 

, 195 (2d Cir.
2014) (granting the BIA’s interpretation in M-E-V-G-
Chevron deference); Chavez v. Att’y Gen., 571 F. App’x 861,
864-65 (11th Cir. 2014) (applying criteria from M-E-V-G-).




                              37
see Gutierrez-Brizuela v. Lynch, 

, 1149 (10th
Cir. 2016) (Gorsuch, J., concurring) (“Chevron and Brand X
permit executive bureaucracies to swallow huge amounts of
core judicial and legislative power and concentrate federal
power in a way that seems more than a little difficult to
square with the Constitution of the framers’ design.”), but it is
the law, and it allows the BIA to change its statutory
interpretation and still be entitled to full deference from
Article III courts, see Nat’l Cable & Telecomms. Ass’n v.
Brand X Internet Servs., 

, 981 (2005) (“Agency
inconsistency is not a basis for declining to analyze the
agency’s interpretation under the Chevron framework.”). All
that is required is that the agency provide a “reasoned
explanation” for its interpretation. 

The BIA has
done so here, and because the three-part test endorsed in
M-E-V-G- is based on a “reasonable construction of the
statute, whether or not it is the only possible interpretation or
… the one [we] might think best,” that test prevails.21 Holder
v. Martinez Gutierrez, 

, 591 (2012).

       21
          At the same time, we are mindful of the role that
courts can and must play to ensure that agencies comply with
their “obligation to render consistent opinions,” Chisholm v.
Def. Logistics Agency, 

, 47 (3d Cir. 1981),
including, as relevant here, review of BIA decisions for
inconsistent application of M-E-V-G’s requirements to
similarly situated petitioners, routine rejection of proposed
PSGs without reasoned explanation, and the imposition of
insurmountable evidentiary burdens that would render
illusory the opportunity to establish a PSG. However, just as
we will carefully examine cases on petition for review to
guard against such dangers, we anticipate that the BIA will
scrutinize the IJ decisions that come before it with those




                               38
             D.     S.E.R.L.    Has     Not      Established
                    Membership       in   a      Cognizable
                    Particular Social Group

        Having concluded that the BIA’s interpretation is
entitled to Chevron deference, we now consider S.E.R.L.’s
claim that her proposed particular social group – immediate
family members of Honduran women unable to leave a
domestic relationship – nevertheless satisfies the test from
M-E-V-G-. To prevail on her asylum and withholding of
removal claims, S.E.R.L. bore the burden of both alleging a
cognizable particular social group as well as establishing her
membership in that group based on evidence of record.
Although “[t]he BIA is not permitted simply to ignore or
misconstrue evidence” in the record, Espinosa-Cortez v. Att’y
Gen., 

, 114 (3d Cir. 2010), we may only reverse
factual findings if we conclude that “the evidence ‘compels’ a
different result.” 

(quoting Elias-

). We agree with the BIA’s
conclusion that S.E.R.L. has not satisfied the social
distinction requirement.

       S.E.R.L. focuses on the legal aspect of our inquiry,
arguing that her proposed social group must be cognizable
because it comprises two groups that the BIA has already
recognized as meeting the particularity and social distinction
requirements: “women of a particular nationality who are
trapped in abusive relationships,” and “immediate family.”
(Reply Br. at 1.) She illustrates her argument by way of a
Venn diagram, suggesting that her group constitutes a


considerations in mind and with an eye towards providing
clear guidance and a coherent body of law in this area.




                             39
particular social group as a matter of logic. While her
reasoning has some superficial appeal, it is flawed, and we
reject it for two reasons.

        First, and most fundamentally, it ignores the factual
feature in determining whether a particular social group is
cognizable. The BIA has repeatedly stated that the particular
social group determination depends on the facts of the case at
hand. See Acosta, 19 I. & N. Dec. at 233-34 (“The particular
kind of group characteristic that will qualify under this
construction remains to be determined on a case-by-case
basis.”); accord Matter of L-E-A-, 27 I. & N. Dec. 40, 42
(BIA 2017); M-E-V-G-, 26 I. & N. Dec. at 251. And that
must naturally be so, once it is given that social distinction
involves proof of societal views. What those views are and
how they may differ from one society to another are questions
of fact upon which the ultimate legal question of cognizability
rests. Consequently, it does not follow that because the BIA
has accepted that one society recognizes a particular group as
distinct that all societies must be seen as recognizing such a
group. Kinship, marital status, and domestic relationships can
each be a defining characteristic of a particular social group,
but that does not mean that adding two or more of those
characteristics together necessarily establishes a cognizable
particular social group. In fact, that kind of addition may well
broaden, rather than narrow, a group such that the society in
question would not recognize it as distinct. Thus, as a matter
of logic, it is invalid to assert that proof in one context is
proof in all contexts.

       Second and closely related, the Board made an
important factual distinction between this case and its prior
decision in Matter of A-R-C-G-. S.E.R.L. relies heavily on




                              40
that decision, in which the Board considered a group
consisting of married female victims of domestic violence.
Matter of A-R-C-G-, 26 I. & N. Dec. 388, 390-95 (BIA 2014),
overruled by Matter of A-B-, 27 I. & N. Dec. 316 (A.G.
2018). Importantly, however, A-R-C-G- was premised on
“DHS’s concession that a particular social group exist[ed],”
based on “unrebutted evidence that Guatemala has a culture
of ‘machismo and family violence.’” 

(citation
omitted). And, as earlier noted, 

n.15, A-R-C-G- has
recently been abrogated by the Attorney General, who stated
that it “caused confusion because it recognized an expansive
new category of particular social groups based on private
violence.” Matter of A-B-, 27 I. & N. Dec. at 319.

       Here, relying on M-E-V-G- and A-R-C-G-, the BIA
concluded that S.E.R.L.’s proposed group failed, in part
because she had not identified sufficient evidence that
immediate family members of Honduran women unable to
leave a domestic relationship are viewed as socially distinct
within Honduran society.

       S.E.R.L. argues that the Board’s decision is
indefensible, because the record parallels what was presented
in A-R-C-G-. She points to evidence, including country
reports documenting violence against Honduran women,
Honduran laws enacted to protect women and victims of
domestic abuse, and evidence suggesting that those laws are
underenforced, as well as a Honduran initiative to combat
violence against women. But that evidence does not compel
the conclusion that S.E.R.L.’s broader proposed group, which
encompasses family members of domestic abuse victims –
including family members who are male or female, young or




                             41
old, and live with or apart from the victims – is socially
distinct.

       We do not read the BIA’s opinion as, in effect,
“ignor[ing] S.E.R.L.’s evidence of rampant violence against
women and their families in Honduras.” (Reply Br. at 17,
18.) To be sure, the record includes disturbing evidence of
crime, gang-related violence, and general human rights
abuses, including gender-based violence against women in
Honduras. The Board, however, noted the lack of evidence in
the record establishing that “members of [S.E.R.L.’s
proposed] group would be perceived, considered, or
recognized in Honduras as a distinct group[.]” (AR at 5.)
Although arguing that the BIA should not be free to credit or
ignore evidence or avoid analyzing precedent just by claiming
that the issue before it is different, S.E.R.L. fails to direct us
to anything in the record that the IJ or BIA has ignored and
that would compel the conclusion that Honduran society
perceives immediate family members of women who cannot
leave domestic relationships as constituting a socially distinct
group. Thus, even if such a group were still cognizable after
the Attorney General’s recent decision overruling A-R-C-G-,
the argument for granting the petition for review in this case
fails.

        S.E.R.L.’s criticism of the BIA’s analysis strikes at the
heart of the Board’s discretion to adopt additional
requirements for identifying a particular social group and its
ability to apply those requirements on a case-by-case basis.
That criticism may or may not be valid but, in any event,
should be directed to Congress. As the law stands now, the
BIA has the discretion it exercised, and while it remains to be
seen whether the application of those requirements proves




                               42
principled and consistent, what matters for our purposes is
that they are capable of such application. Martinez 

, 596. In light of the deference owed to the
BIA’s view of the INA, and after reviewing the record as a
whole, we conclude that S.E.R.L. has not met her burden of
showing that the evidence here compels the conclusion that
her proposed social group is viewed in Honduras as being
socially distinct.22

III.   CONCLUSION

       For the reasons set forth, we will deny S.E.R.L.’s
petition for review.23


       22
          Because we agree that S.E.R.L. has not adduced
sufficient evidence to establish the existence of her proposed
particular social group, we do not reach any of the other bases
for the BIA’s denial of her application for asylum and
withholding of removal, and her final argument suggesting
that remand is required for the agency to address her well-
founded fear of persecution by Juan Orellana is moot.
       23
           S.E.R.L.’s outstanding motion to supplement the
record will also be denied. That motion is premised on a
motion to reopen proceedings before the BIA. The BIA
denied her motion, and S.E.R.L. has not appealed that
decision. Thus, her motion and the new evidence it discusses
are not properly before us. See Kamara v. Att’y Gen., 

, 218 (3d Cir. 2005) (noting that we “must approve
or reject the agency’s action purely on the basis of the reasons
offered by, and the record compiled before, the agency itself”
(citation omitted)).




                              43