Schafer v. Astrue

PUBLISHED

UNITED STATES COURT OF APPEALS
FOR THE FOURTH CIRCUIT

JANICE SCHAFER, Guardian Ad 
Litem for WMS, infant,
Plaintiff-Appellant,
v.  No. 10-1500

MICHAEL J. ASTRUE,
Defendant-Appellee.

Appeal from the United States District Court
for the Western District of North Carolina, at Charlotte.
Graham C. Mullen, Senior District Judge.
(3:09-cv-00096-GCM-DSC)

Argued: January 28, 2011

Decided: April 12, 2011

Before WILKINSON, AGEE, and DAVIS, Circuit Judges.

Affirmed by published opinion. Judge Wilkinson wrote the
majority opinion, in which Judge Agee joined. Judge Davis
wrote a dissenting opinion.

COUNSEL

ARGUED: Kelsi Brown Corkran, UNITED STATES
DEPARTMENT OF JUSTICE, Washington, D.C., for Appel-
lee. ON BRIEF: George L. Fitzgerald, Charlotte, North Caro-
2 SCHAFER v. ASTRUE
lina, for Appellant. Tony West, Assistant Attorney General,
William Kanter, UNITED STATES DEPARTMENT OF
JUSTICE, Washington, D.C.; Anne M. Tompkins, United
States Attorney, Charlotte, North Carolina; David Black, Gen-
eral Counsel, Karen Aviles, General Attorney, SOCIAL
SECURITY ADMINISTRATION, Baltimore, Maryland, for
Appellee.

OPINION

WILKINSON, Circuit Judge:

Don and Janice Schafer married in 1992. Don died the next
year. With the help of in vitro fertilization, however, Janice
gave birth to W.M.S., Don Schafer’s biological child, a num-
ber of years later. Janice Schafer then applied on W.M.S.’s
behalf for survivorship benefits under the Social Security Act.
See 42 U.S.C. § 402(d) et seq.

The Social Security Administration rejected W.M.S.’s
claim. Because under its view natural children must be able
to inherit from the decedent under state intestacy law or sat-
isfy certain exceptions to that requirement in order to count as
"children" under the Act, W.M.S. was not eligible for survi-
vorship benefits. See 42 U.S.C. §§ 416(h)(2), (h)(3)(C). The
district court agreed. On appeal, Schafer contends that undis-
puted natural children such as W.M.S. plainly fall within 42
U.S.C. § 416(e)(1)’s basic definition of "child," making their
state intestacy rights irrelevant.

We shall affirm the judgment. The agency’s view best
reflects the statute’s text, structure, and aim of providing ben-
efits primarily to those who unexpectedly lose a wage earn-
er’s support. And even if the agency’s interpretation were not
the only reasonable one, it falls well within the range of per-
missible readings entitled to deference under Chevron U.S.A.
 SCHAFER v. ASTRUE 3
Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837
(1984).

I.

The sad facts giving rise to this case are not in real dispute.
Janice and Don Schafer, Jr. were married in June 1992. Four
months later he was diagnosed with cancer and informed that
the chemotherapy he needed might render him sterile. In
December 1992 he deposited sperm samples with a long-term
storage facility, but in March 1993 he died of a heart attack.
At the time Don was domiciled in Virginia.

In April 1999 Janice Schafer conceived a child through in
vitro fertilization, and she gave birth to that child, W.M.S., on
January 13, 2000 in Texas.1 There is significant evidence that
W.M.S. is Don Schafer’s biological child, born almost seven
years after Don’s death. There is also evidence that Don
intended for Janice to use the stored semen to conceive a child
after his anticipated death, though he never expressed consent
in writing to be the legal father of a child resulting from post-
humous in vitro fertilization. In 2001 a Texas court purported
to declare Don Schafer W.M.S.’s father.

In 2004 Janice Schafer applied on W.M.S.’s behalf to the
Social Security Administration ("SSA") for surviving child
benefits under the Social Security Act ("the Act"), see 42
U.S.C. § 402(d). An administrative law judge initially
awarded W.M.S. benefits, but the SSA’s Appeals Council
reversed, reasoning that W.M.S. was not Don Schafer’s
"child" within the meaning of the Act because W.M.S. could
1
Although the parties refer to W.M.S. as being conceived through artifi-
cial insemination in their briefs, claimant’s application for benefits states
that he was conceived through in vitro fertilization. While the two proce-
dures are medically distinct, the specific type of assistive reproductive
technology utilized is immaterial to our analysis in this case. We therefore
use the terminology from the application.
4 SCHAFER v. ASTRUE
not inherit from him under Virginia intestacy law. Janice
Schafer then sought review of the SSA’s decision in federal
district court, which upheld the SSA’s denial of benefits. She
now appeals.

II.

A.

Every child claiming survivorship benefits under the Act
must meet a series of requirements. Initially, the child or his
guardian must have filed an application. 42 U.S.C.
§ 402(d)(1)(A). The child must also fit certain substantive
criteria: he must be unmarried and either under certain age
limits or subject to a disability, id. § 402(d)(1)(B), and he
must have been "dependent upon such individual . . . at the
time of such [individual’s] death . . . ," id. § 402(d)(1)(C)(ii)
(internal subsection division omitted).2

Before even reaching these questions, however, an appli-
cant must establish something more fundamental: that he is
the insured’s "child" within the meaning of the Act. The Act’s
basic grant of benefits provides that "[e]very child (as defined
in section 416(e) of this title) . . . of an individual who dies
a fully or currently insured individual . . . shall be entitled to
a child’s insurance benefit . . . ." Id. § 402(d)(1). As relevant
here, § 416(e) is sparse: "The term ‘child’ means (1) the child
or legally adopted child of an individual." Id. § 416(e)(1).

Section 416(e)(1), however, is not the only provision of the
Act that bears on the determination of child status. Section
416, titled "Additional definitions," also includes § 416(h),
labeled "Determination of family status." That provision
states:
2
There is some question whether W.M.S. could satisfy
§ 402(d)(1)(C)(ii)’s dependency requirement, but because we agree with
the SSA that W.M.S. is not Don Schafer’s "child" under the Act we need
not resolve this issue.
 SCHAFER v. ASTRUE 5
In determining whether an applicant is the child . . .
of a fully or currently insured individual for purposes
of this subchapter, the Secretary shall apply such law
as would be applied in determining the devolution of
intestate personal property by the courts of the State
in which such insured individual . . . was domiciled
at the time of his death . . . . Applicants who accord-
ing to such law would have the same status relative
to taking intestate personal property as a child . . .
shall be deemed such.

Id. § 416(h)(2)(A).

Section 416(h) also provides three additional gateways to
child status for those who cannot establish it through
§ 416(h)(2)(A)’s intestacy provision. First, an applicant who
"is a son or daughter of a fully or currently insured individ-
ual" but cannot inherit from that individual is deemed a child
if his parents went through a marriage ceremony that turned
out to be legally invalid. Id. § 416(h)(2)(B). Second, a child
who cannot inherit from a deceased insured individual under
state intestacy law is a "child" under the Act where prior to
death the decedent "had acknowledged [parentage] in writ-
ing," "had been decreed [the child’s parent] by a court," or
"had been ordered by a court to contribute to the support of
the applicant because the applicant was [the insured individu-
al’s child]." Id. § 416(h)(3)(C)(i)(I)–(III) (internal subsection
divisions omitted). Third, a child who cannot inherit is
deemed a "child" if "such insured individual is shown by evi-
dence satisfactory to the Secretary to have been the [parent]
of the applicant, and such insured individual was living with
or contributing to the support of the applicant at the time such
insured individual died." Id. § 416(h)(3)(C)(ii).

In addition, the Act gives the Commissioner of Social
Security rulemaking authority. See id. § 405(a).
6 SCHAFER v. ASTRUE
B.

This case turns on the relationship between the brief defini-
tion of "child" in § 416(e)(1) — which is part of the only defi-
nition referred to in § 402(d)(1)’s basic grant of benefits—and
§ 416(h)’s more specific provisions. On the SSA’s view,
§ 416(h) "provides the analytical framework that [it] must fol-
low for determining whether a child is the insured’s child" for
purposes of § 416(e)(1)’s definition. Social Security Acquies-
cence Ruling 05-1(9), 70 Fed. Reg. 55,656, 55,657 (Sept. 22,
2005) ("Acquiescence Ruling"); see also 20 C.F.R.
§ 404.355(a) (using § 416(h)’s provisions as the only means
of establishing child status as a natural child). In keeping with
this view, the SSA has always required applicants claiming
natural child status—including the undisputed biological chil-
dren of married parents—to pass through one of § 416(h)’s
pathways to secure that status. See, e.g., Old-Age, Disability,
Dependents’ and Survivors’ Insurance Benefits, 44 Fed. Reg.
34,479, 34,487 (June 15, 1979) (listing the same requirements
in an earlier version of 20 C.F.R. § 404.355 as part of a
restatement of the applicable regulations).

The SSA has also taken the view that posthumously con-
ceived children such as W.M.S. can qualify as "children"
under the Act only through the state intestacy provision. See
Acquiescence Ruling, 70 Fed. Reg. at 55,657 ("[T]o meet the
definition of ‘child’ under the Act, an after-conceived child
must be able to inherit under State law."); Program Opera-
tions Manual System GN 00306.001(C)(1)(c), available at
https://secure.ssa.gov/poms.nsf/lnx/0200306001 (same).3 The
insured parent of such a child by definition died prior to the
3
Although Schafer’s primary argument is that W.M.S.’s inheritance
rights are irrelevant to child status, she argues alternatively that W.M.S.
can inherit from Don Schafer under Virginia law. Because she does not
contend W.M.S. has child status under any of § 416(h)’s other provisions,
we need not address the SSA’s position that posthumously conceived chil-
dren may only secure child status through § 416(h)(2)(A).
 SCHAFER v. ASTRUE 7
child’s conception, and therefore parentage could not have
been acknowledged or decreed prior to death, nor could the
applicant have been living with or receiving contributions
from the decedent when the decedent passed away.

On the SSA’s view, then, W.M.S. is entitled to benefits
only if he could inherit from Don Schafer under Virginia law.
But Virginia law does not recognize any "child born more
than ten months after the death of a parent" as that parent’s
child for intestacy purposes. Va. Code Ann. § 20-164 (ten-
month limitation); Va. Code Ann. § 64.1-5.1(2) (incorporat-
ing the provisions of Va. Code Ann. § 20-156 et seq. in deter-
mining the parentage of "a child resulting from assisted
conception"). W.M.S., however, was born almost seven years
after Don Schafer’s death. The SSA therefore denied
W.M.S.’s claim.

Schafer takes a very different view of the relationship
between § 416(e)(1) and § 416(h). Adopting the view of
Gillett-Netting v. Barnhart, 371 F.3d 593, 597 (9th Cir. 2004),
she argues that § 416(h)’s strictures do not apply to children
whose "parentage . . . is not disputed." After all, § 402(d)
refers explicitly to § 416(e), not § 416(h), in defining "child,"
and on her view § 416(e)(1)’s plain terms — "the child or
legally adopted child of such individual" — obviously cover
such children. See Weinberger v. Salfi, 422 U.S. 749, 781
n.12 (1975) (describing § 416(e)(1)’s reference to a "child" as
a reference to a "natural" child). It follows, she argues, that
W.M.S.’s ability to inherit from Don Schafer under Virginia
intestacy law is irrelevant to determining W.M.S.’s child sta-
tus. Instead, as Don and Janice Schafer’s undisputed biologi-
cal child, W.M.S. must only show that he meets the other
statutory conditions described above. See 42 U.S.C.
§ 402(d)(1)(A)–(C). The Third Circuit has recently joined
Gillett-Netting in accepting Schafer’s reading in posthumous
conception cases, see Capato ex rel. B.N.C. v. Comm’r of Soc.
Sec., ___ F.3d ___, 2011 WL 9368 (3d Cir. Jan. 4, 2011), and
the Eighth Circuit is currently facing the question, see Beeler
8 SCHAFER v. ASTRUE
v. Astrue, No. C09-0019 (N.D. Iowa Nov. 12, 2009), appeal
docketed, No. 10-1092.

III.

In deciding between these warring interpretations, it is
important to bear in mind the rules of engagement. We first
ask, at Chevron step one, whether "Congress has directly spo-
ken to the precise question at issue." Chevron, 467 U.S. at
842. If so, the agency is not free to counter Congress’s com-
mand, and the courts must uphold faithful interpretations and
reject disobedient ones. But where "a statute is silent or
ambiguous with respect to the specific issue," a reviewing
court at Chevron step two asks "whether the agency’s answer
is based on a permissible construction of the statute." Id. at
843.

At first blush, Schafer’s proposed course—rejecting the
SSA’s interpretation at Chevron step one—is an alluring one.
After all, "the plain language of the statute" is "the most reli-
able indicator of Congressional intent," Soliman v. Gonzales,
419 F.3d 276, 281–82 (4th Cir. 2005), and the plain meaning
of "child" in § 416(e)(1) might seem necessarily to include the
biological children of those who, because of tragic circum-
stances, could only become parents after their death. But for
many reasons, we think this approach would be unsound. The
"traditional tools of statutory construction," Chevron, 467
U.S. at 843 n.9, demonstrate that Congress intended the SSA
to use state intestacy law, and even if it did not precisely
speak to the question, the SSA’s reasonable interpretation is
entitled to deference.

A.

We begin by examining the text of the Act. Section 416(e)
itself is notably brief: "The term ‘child’ means (1) the child
or legally adopted child of an individual . . . ." Although the
"child" portion of this language does aim at natural children,
 SCHAFER v. ASTRUE 9
see Salfi, 422 U.S. at 781 n.12, other courts have noted that
"this definitional tautology . . . does not provide much guid-
ance" as to how the SSA should go about making that child
status determination. Conlon ex rel. Conlon v. Heckler, 719
F.2d 788, 800 (5th Cir. 1983).

This lack of guidance highlights the difficulty we have with
the position of our colleagues on the Third and Ninth Circuits.
Those courts have been willing to elevate the sparse definition
found in § 416(e)(1) and to completely de-emphasize the
more extensive definition found in § 416(h)(2)(A), thereby
treating Congress’s more comprehensive efforts as a mere
afterthought. In so doing, they have overlooked Congress’s
plain and explicit instruction on how the determination of
child status should be made: "In determining whether an
applicant is the child . . . of a fully or currently insured indi-
vidual for purposes of this subchapter, the Secretary shall
apply such law as would be applied in determining the devo-
lution of intestate personal property . . . ." 42 U.S.C.
§ 416(h)(2)(A) (emphasis added). The text in short could
hardly be more clear.

It would be startling if Congress had failed to provide
greater guidance on child status than that set forth in
§ 416(e)(1). The term "child" lies at the core of the Act’s ben-
efit program, and until recent years proving biological parent-
age scientifically was next to impossible. See Jill Handley
Andersen, The Functioning Father: A Unified Approach to
Paternity Determinations, 30 J. Fam. L. 847, 852-53 & nn.36-
45 (1991-92) (tracing the evolution of biological paternity
testing). And though modern technology makes determining
biological parentage scientifically feasible, it also creates situ-
ations that illustrate the complexity of determining the mean-
ing of "child" today. See Capato, 2011 WL 9368, at *5.
Imagine, for instance, a child born after one woman’s fertil-
ized ovum was implanted in and carried to term by another
woman. It is unclear, as a linguistic matter, whether the child
is the birth mother’s "child," the biological mother’s "child,"
10 SCHAFER v. ASTRUE
or, somehow, both. In short, we think it very unlikely Con-
gress would have left the SSA so utterly in the dark about
such a critical term.

Of course, Capato and Gillett-Netting contend that
§ 416(e)(1) provides plenty of guidance, insisting that because
"child" ordinarily means "natural child," § 416(e)(1) indepen-
dently provides child status to those children whose "natural,
or biological" parentage is not "disputed." Gillett-Netting, 371
F.3d at 596, 597. But this cannot be right. To begin with, it
would attribute inconsistent views about child status to the
Congress. In § 416(h)(3)(C)(ii), Congress provided child sta-
tus to a child who cannot inherit but who can prove both that
the insured was the child’s parent and that the insured was
"living with or contributing to" the child at the time of death.
But if undisputed biological parentage is enough under
§ 416(e)(1), it would have made no sense for Congress to
require those whose parentage was initially disputed but was
then resolved to prove something in addition to biological
parentage. Similarly, § 416(h)(2)(B) grants child status to a
non-inheriting child only if he or she "is the son or daughter
of [the] insured" — demonstrating that parentage is no longer
in dispute — and if his or her parents went through a legally
invalid marriage ceremony. Congress would not have
imposed an additional proof requirement on these undisputed
children if undisputed biological parentage sufficed under
§ 416(e)(1).

Sensing this difficulty of reading "child" to mean natural
child, Capato and Gillett-Netting alternatively hint that the
class of children independently provided child status by
§ 416(e)(1) is comprised of the undisputed biological children
of married parents. See Capato, 2011 WL 9368, at *3; Gillett-
Netting, 371 F.3d at 596. This approach solves part of the
problem by "explaining" why out-of-wedlock children strug-
gled to gain child status, but the solution comes at too high
a price. Both Gillett-Netting and Capato insist that the plain
English meaning of "child" in § 416(e)(1) drives their views.
 SCHAFER v. ASTRUE 11
See Capato, 2011 WL 9368, at *3; Gillett-Netting, 371 F.3d
at 596. But whatever their legal rights may be, out-of-wedlock
children are indisputably the natural children of their biologi-
cal parents under the ordinary English usage of the term.
Therefore, Capato and Gillett-Netting cannot state that
§ 416(e)(1) covers the children of only married couples with-
out contradicting their earlier claim that § 416(e)(1)’s suppos-
edly obvious usage of "child" meant all natural children.

It is clear, then, that § 416(e)(1) cannot provide all of the
answers. But there is no need for us to join Capato and
Gillett-Netting in crafting from whole cloth § 416(e)(1)’s
meaning without reference to the Act’s other provisions when
the plain text of those provisions offers all of the guidance
that is needed. Section 416(h)(2)(A) describes precisely how
one ought to determine the meaning of "child" under
§ 416(e)(1): "In determining whether an applicant is the child
. . . of a fully or currently insured individual for purposes of
this subchapter" — that is, for the purposes of federal old-
age, survivors, and disability insurance benefits — "the Secre-
tary shall apply such [state] law as would be applied in deter-
mining the devolution of intestate personal property . . . ." 42
U.S.C. § 416(h)(2)(A) (emphasis added).

This language contains no textual suggestion that it is lim-
ited to disputed or out-of-wedlock children. Indeed, every-
thing about it suggests the opposite: it speaks of applying state
intestacy law for purposes of the whole Act rather than for
purposes of determining child status in disputed parentage
cases, and it specifically addresses itself to the child status
determination that must take place in evaluating every bene-
fits application.

Of course, the SSA’s interpretation might seem to create a
few textual puzzles of its own. If Congress intended all appli-
cants to proceed through one of § 416(h)’s channels in claim-
ing child status under § 416(e)(1), it could have referred to
that provision rather than § 416(e) in defining "child" for pur-
12 SCHAFER v. ASTRUE
poses of § 402(d)’s basic grant of benefits. And if § 416(e)(1)
does not independently provide child status to even the undis-
puted biological child of married parents, Congress may have
been better off leaving it out of the statute, an outcome the
SSA’s interpretation might seem to reach.

These difficulties are more apparent than real. Consider
first the point about Congress’s reference to § 416(e), not
§ 416(h), in § 402(d)’s basic grant. As an initial matter, it is
not at all unusual for Congress to refer explicitly only to one
section even though some of that section’s terms are given
their full import by another, unmentioned section. See, e.g, 29
U.S.C. § 626(a) (granting the EEOC investigative authority
"in accordance with the powers and procedures provided" in
29 U.S.C. §§ 209 and 211, the first of which largely just
incorporates 15 U.S.C. §§ 49 and 50).

More importantly, on the SSA’s view § 416(e) still has
work to do even if § 416(h) provides the ultimate framework
for securing child status under § 416(e)(1). Section 416(e)(1)
clarifies that natural children and legally adopted children are
included, § 416(e)(2) includes certain stepchildren, and
§ 416(e)(3) includes some grandchildren and stepgrandchil-
dren. See Conlon, 719 F.2d at 800 ("Section 416(e) primarily
places stepchildren and certain grandchildren within the defi-
nition of ‘child.’"). This inclusion importantly expands the
scope of the Act and distinguishes it from narrower benefits
programs. In Cleland v. Office of Personnel Management, 984
F.2d 1193, 1195 (Fed. Cir. 1993), for example, the court
rejected a federal employee’s grandchild’s claim to survivor-
ship benefits under 5 U.S.C. § 8341(e)(2) and referred to
§ 416(e) in noting that "[w]hen Congress wishes to include
grandchildren within the meaning of ‘child’ for purposes of a
statute it knows how to do so." Thus, even if § 416(h) pulls
the laboring oar in defining "child" within the context of natu-
ral children, § 416(e) still plays an important role.

Nor is it particularly significant that § 416(e)(1) itself plays
a complementary role on SSA’s view. For one thing, it still
 SCHAFER v. ASTRUE 13
contributes to § 416(e)’s overall purpose — cataloging the
kinds of children included as "children" under the Act—by
specifying that natural children and legally adopted children
are included. For another, it is not clear just how the SSA
could give "full meaning" to the statutory proposition that "a
‘child’ is a child" without help from neighboring provisions,
especially once one recognizes that giving it too exclusively
biological a meaning conflicts with § 416(h)(3)(C)(ii).

Considering all of the textual evidence, then, we think Con-
gress plainly intended the SSA to apply the Act as the SSA
always has. The SSA’s view follows to the letter Congress’s
explicit and precise instruction as to how the agency should
determine child status under the Act. Moreover, the SSA’s
interpretation, rather than focusing myopically on a single
term, makes sense of the statute as a whole, see Soliman, 419
F.3d at 282, and recognizes that the "meaning . . . of certain
words or phrases may only become evident" when considered
in textual and historical context, FDA v. Brown & Williamson
Tobacco Corp., 529 U.S. 120, 132 (2000). Whether one looks
at the instruction of § 416(h)(2)(A) individually or at the stat-
ute as a whole, the SSA’s view is dutiful and faithful to Con-
gress’s intent.

B.

The Act’s history also suggests that the SSA is respecting
Congress’s wishes. See, e.g., Elm Grove Coal Co. v. Dir.,
O.W.C.P., 480 F.3d 278, 293-94 (4th Cir. 2007) (consulting
the history of Congress’s efforts at Chevron step one). We are
of course not the first court to look to the Act’s history in
determining its meaning. Gillett-Netting relied heavily on its
view that § 416(h)(2) and § 416(h)(3) "were added to the Act
to provide various ways in which children could be entitled to
benefits even if their parents were not married or their parent-
age was in dispute." Gillett-Netting, 371 F.3d at 596. As a
result, Gillett-Netting believed those provisions have no appli-
cation to those who do not face such circumstances.
14 SCHAFER v. ASTRUE
But insofar as Gillett-Netting meant to suggest the intestacy
provision was not part of the original statutory framework, it
is simply wrong. The original version of the Act contained a
framework identical in fundamentals to the one on the books
today: a grant of benefits to every "child" as defined by one
provision of the Act, see Social Security Act Amendments of
1939, Pub. L. No. 76-379, § 202(c)(1), 53 Stat. 1360, 1364;
a definition of "child" in that section as "the child of an indi-
vidual," id. § 209(k), 53 Stat. at 1377; and another definition
instructing the SSA to "apply such law as would be applied
in determining the devolution of intestate personal property"
under state law to "determin[e] whether an applicant is the . . .
child [of the insured]," id. § 209(m), 53 Stat. at 1378.
Although these provisions were altered and renumbered over
the years, the basic structure remained intact. That is, some
version of § 416(h)(2)(A) has always been part of the statu-
tory scheme, weakening any inference that it is just a new
way for disputed children to gain child status.

Indeed, when considered in greater detail, the Act’s legisla-
tive history demonstrates that Congress understood
§ 416(h)(2)(A)’s intestacy provisions to be the backbone of
all child status determinations. The Senate Finance Commit-
tee’s Report to the 1965 Amendments, which added
§ 416(h)(3)(C), stated that "[u]nder present law, whether a
child meets the definition of a child for the purpose of getting
child’s insurance benefits . . . depends on the laws applied in
determining the devolution of intestate personal property
. . . ." S. Rep. No. 89-404, pt. 1, at 109 (1965), reprinted in
1965 U.S.C.C.A.N. 1943, 2049.

Congress’s belief in this regard was well-grounded: a great
deal of case law held that all those claiming child status had
to prove the ability to inherit under state law. See, e.g., Gainey
v. Flemming, 279 F.2d 56 (10th Cir. 1960) (affirming denial
of benefits to undisputed children of an invalid marriage
because the children could not inherit); Robles v. Folsom, 239
F.2d 562 (2d Cir. 1956) (Hand, J.) (affirming denial of bene-
 SCHAFER v. ASTRUE 15
fits to an undisputed child who had always been supported by
his unmarried father because the child could not inherit);
Hobby v. Burke, 227 F.2d 932, 933 n.3 (5th Cir. 1956) (noting
that § 416(e)’s definition of "child" is "further defined" in
§ 416(h)’s intestacy provisions).

The 1965 Amendments themselves demonstrate Congress’s
continued acceptance of this basic framework. The Senate
Report noted the problem the framework created: children
whom Congress viewed as similarly situated were treated dif-
ferently for child status purposes simply because state law
denied out-of-wedlock children inheritance rights. See S. Rep.
No. 89-404, 1965 U.S.C.C.A.N. 1943, at 2050. To fix the
problem, however, Congress did not change the received view
of the law and "clarify" that all undisputed children are "chil-
dren" under § 416(e)(1). Instead, it worked within the frame-
work of the background intestacy provision, adding
§ 416(h)(3)(C) so that child status could also exist where par-
entage was acknowledged, decreed, implicit in a contribution
order, or proven along with cohabitation or contribution. See
id.; McMillian ex rel. McMillian v. Heckler, 759 F.2d 1147,
1149–50 (4th Cir. 1985) (noting that § 416(h)(3)(C) provides
means for out-of-wedlock children to establish child status).

Indeed, § 416(h)(3) certainly sounds like an exception to a
general requirement: "An applicant who is the son or daughter
of a fully . . . insured individual, but who is not (and is not
deemed to be) the child of such insured individual under [the
state intestacy or legally invalid marriage provisions], shall
nevertheless be deemed to be the child of such insured indi-
vidual" upon meeting one of § 416(h)(3)(C)’s conditions. 42
U.S.C. § 416(h)(3) (emphasis added).

The Act’s legislative history could hardly be clearer. Con-
gress understood the Act’s framework as requiring all natural
children to pass through § 416(h) to claim child status, and it
drafted in keeping with this view in adding § 416(h)(3)(C).
16 SCHAFER v. ASTRUE
C.

The SSA’s interpretation also best comports with the Act’s
purposes. See, e.g., Elm Grove, 480 F.3d at 293 (looking to
the statute’s purpose at Chevron step one). The Act is not a
"welfare program generally benefiting needy persons," but
rather an effort to "provide the . . . dependent members of [a
wage earner’s] family with protection against the hardship
occasioned by his loss of earnings." Califano v. Jobst, 434
U.S. 47, 52 (1977); see also S. Rep. 89-404, 1965
U.S.C.C.A.N. 1943, at 2050 (describing survivorship benefits
as "a national program . . . intended to pay benefits to replace
the support lost by a child when his father . . . dies . . . .").
The Act thus creates a core beneficiary class: the children of
deceased wage earners who relied on those earners for sup-
port.

The Act’s child status provisions reflect this basic aim. For
instance, § 416(h)(3)(C)(ii) grants child status to those who
are proven to be the biological child of the insured provided
the insured was "living with or contributing to" the child at
the time of death. Similarly, § 416(h)(3)(C)(i), which grants
child status to those whose parentage is acknowledged,
decreed, or implicit in a contribution order, suggests similar
dependence on the wage earner because the acknowledgment
or decree must occur prior to the insured’s death and because
such acknowledgment or decree is likely to be accompanied
by financial outlays from the insured. Section 416(h)(2)(A) is
no different. As the Supreme Court has noted in assessing the
Act’s deemed dependency provisions, "where state intestacy
law provides that a child may take personal property from a
father’s estate, it may reasonably be thought that the child will
more likely be dependent during the parent’s life and at his
death." Mathews v. Lucas, 427 U.S. 495, 514 (1976).

The SSA’s interpretation, by tying natural child status
determinations to one of § 416(h)’s pathways, thus reflects the
Act’s basic aim of primarily helping those children who lost
 SCHAFER v. ASTRUE 17
support after the unanticipated death of a parent. Posthu-
mously conceived children, however, differ from members of
the core beneficiary class in two ways. First, they necessarily
could not have relied on the wage earner’s wages prior to his
death. Second, they generally come into being after it is clear
that one of the parents will not be able to support the child in
the ordinary way during the child’s lifetime, meaning that the
survivorship benefits would serve a purpose more akin to sub-
sidizing the continuance of reproductive plans than to insuring
against unexpected losses.

These differences might have persuaded Congress to
exclude most posthumously conceived children from child
status under an updated version of the Act. But the current
statute, on the SSA’s interpretation, properly includes as chil-
dren those posthumously conceived children whom state law-
makers conclude are similarly situated enough to more
traditionally conceived children that they deserve a share in
the decedent’s estate. In making this determination, state post-
humous conception schemes often reflect the same concerns
animating the Act’s creation of the core beneficiary class,
thereby both protecting the Act’s chief intended aid recipients
and expanding child status to cover closely analogous cases.

Through heightened consent requirements, for example,
state posthumous conception laws ensure that child status and
inheritance rights go only to those whom the deceased parent
intended to support as a child. See, e.g., Va. Code Ann. § 20-
158(B) (generally requiring consent in writing to be the parent
of any child resulting from implantation after one’s death);
Colo. Rev. Stat. § 19-4-106(8) (same); Utah Code Ann.
§ 78B-15-707 (same). The time limitations often imposed on
posthumously conceived children’s inheritance rights simi-
larly reflect the concerns underlying the Act by protecting the
estate for already existing children, the core intended benefi-
ciaries of the Act. See, e.g., Va. Code Ann. § 20-164 (ten-
month limit); Cal. Prob. Code § 249.5(c) (two-year limit from
18 SCHAFER v. ASTRUE
acknowledged death); La. Rev. Stat. Ann. § 9:391.1(A)
(three-year limit).

By contrast, Schafer’s interpretation contravenes the stat-
ute’s purpose by threatening the core beneficiary class. She
claims the Act provides child status to any undisputed biolog-
ical child. But, as the Commissioner notes, the Act limits the
total benefits payable from one employment record. See 42
U.S.C. § 403(a)(1). As a result, where an additional child
claims benefits from a record, children already claiming from
it could see a reduction in their benefits. Though the addi-
tional benefits would generally stay in the same family, it
remains true that existing children, the Act’s core intended
beneficiaries, could receive proportionately less support. Con-
gress designed the Act with those children in mind, and the
SSA’s interpretation best protects their interests.

D.

Our conclusion that Congress intended the SSA to use state
intestacy law as it does is buttressed by a good deal of case
law favoring the SSA’s view. For instance, Conlon squarely
held that all child status determinations must pass through
§ 416(h). The court noted that § 416(e)’s primary function is
to include some stepchildren and grandchildren within the
scope of the Act, not to fully define "child," and it pointed out
the difficulty in seeing § 416(e)(1)’s sparse language as the
Act’s main provision dealing with child status. See Conlon,
719 F.2d at 800. The court also reasoned that Congress did
not leave the SSA adrift; "[s]ection 416(e)(1) is . . . modified
by sections 416(h)(2) and (3) which provide the final words
on who is to be considered a child for purposes of section
416(e)(1)." Id. (emphasis added). That is, "because [the appli-
cant did] not fit within the definition of a ‘child’ under either
section 416(h)(2)(A) or section 416(h)(3)(C)(i)(II), she would
not be considered a child under section 416(e)." Id.

Many cases also reflect the SSA’s view. For example,
many of the state supreme courts to consider the inheritance
 SCHAFER v. ASTRUE 19
rights of posthumously conceived children did so only
because a federal district court certified the state law question
to them as dispositive of a benefits claim. See, e.g., Finley v.
Astrue, 270 S.W.3d 849 (Ark. 2008); Khabbaz ex rel. Eng v.
Comm’r, Soc. Sec. Admin., 930 A.2d 1180 (N.H. 2007);
Woodward v. Comm’r of Soc. Sec., 760 N.E.2d 257 (Mass.
2002).

Federal circuit courts have also accepted the SSA’s posi-
tion. Recently the D.C. Circuit noted:

The Act establishes the following procedure to deter-
mine whether an individual qualifies as a "child."
"Child" is defined [in § 416(e)], inter alia, as "the
child or legally adopted child of an individual." To
determine whether an applicant meets the Act’s defi-
nition of "child," the SSA must: "apply such law as
would be applied in determining the devolution of
intestate personal property . . . ."

Javier v. Comm’r of Soc. Sec., 407 F.3d 1244, 1247 (D.C. Cir.
2005) (quoting 42 U.S.C. § 416(h)(2)(A)) (citation omitted).
Similarly, DeSonier v. Sullivan, 906 F.2d 228, 229–30 (6th
Cir. 1990), stated that "[t]he ALJ recognized that a claimant’s
relationship to a deceased wage earner is determined by
applying [state intestacy law]" and that "[d]etermination of a
claimant’s family status is made by applying the provisions of
42 U.S.C. § 416(h)."

None of these cases mention any restriction of § 416(h)’s
operation to cases involving disputed parentage. Although
many involve such circumstances, that fact is more attribut-
able to the paucity of benefit denials where a legitimate bio-
logical child seeks benefits than to judicial disagreement with
the SSA’s long-held view; as the SSA has noted, every state
provides inheritance rights to such children. See Application
of State Law in Determining Child Relationship, 63 Fed. Reg.
57,590, 57,591 (Oct. 28, 1998). We think it unlikely that so
20 SCHAFER v. ASTRUE
many courts over so many years have been wrong, and obvi-
ously so, in their understanding of Congress’s intent.

E.

The text, legislative history, purpose, and prior judicial
approaches to the Act indicate that Congress wanted the SSA
to apply § 416(h) in determining child status. Schafer’s con-
trary view takes a single, sparse statutory word and makes it
the statute’s primary statement on the majority of survivorship
benefit applications. It relegates Congress’s most detailed and
comprehensive effort to define "child" under the Act to the
status of a fallback provision, one that only applies where par-
entage is "disputed." This limitation runs against the grain of
§ 416(h)’s text, § 416(h)(3)(C)’s history, and the SSA’s obvi-
ous need for congressional guidance. This seems a wrong
turn, but fortunately it is not one the statute compels us to
take. Indeed, the Act tells us to go in the other direction, the
one that the SSA has always followed.

IV.

The SSA’s position becomes even stronger considering the
administrative law context in which we find ourselves. Even
if one were to disagree that the Act dictates the SSA’s inter-
pretation, the considerable evidence for the SSA’s view at
least demonstrates statutory ambiguity. To recapitulate,
§ 416(h)(2)(A)’s clear instruction and comprehensive
approach significantly undermines the view that § 402(d)’s
basic grant and § 416(e)(1)’s sparse definition independently
confer child status on undisputed children. Gillett-Netting and
Capato cannot offer an interpretation of § 416(e)(1) that
makes sense of § 416(h)(3)(C). The 1965 Amendments’ his-
tory indicates that Congress drafted them with the understand-
ing that all child status determinations were made initially
through state intestacy law, calling into question Schafer’s
view that § 416(h)’s provisions are obviously just alternate
pathways. And the longstanding, ubiquitous judicial accep-
 SCHAFER v. ASTRUE 21
tance of the SSA’s interpretation suggests it is far from unrea-
sonable.

Given that the Act is at least ambiguous, the barrier to
Schafer’s challenge is steep. "[W]e may not disturb an agency
rule [at Chevron step two] unless it is ‘arbitrary or capricious
in substance, or manifestly contrary to the statute.’" Mayo
Found. for Med. Educ. & Research v. United States, 131 S.
Ct. 704, 711 (2011) (quoting Household Credit Servs., Inc. v.
Pfennig, 541 U.S. 232, 242 (2004)). Chevron deference of this
sort stands at the heart of modern administrative law. It
ensures that agency officials, who are subject to greater politi-
cal accountability and possess greater relevant expertise than
judges, take the lead in implementing programs delegated to
their care. See Chevron, 467 U.S. at 865–66. The SSA’s inter-
pretation easily passes muster under this deferential standard.

Of course, the same factors that led us to conclude above
that the SSA’s interpretation best respects Congress’s intent
demonstrate that the SSA’s resolution of any statutory
ambiguity is a permissible one. But there is more to be said
in favor of deferring to the SSA’s interpretation. Deferring
here affords Congress, the SSA, and the federal courts their
proper places in the modern regulatory scheme.

Congress, of course, has the last word in determining the
scope of federal survivorship benefits. It is unlikely that Con-
gress is unaware of the SSA’s longstanding interpretation of
the Act. Given the rapid rise of issues related to reproductive
technology, Congress would hardly be ignorant of the fact
that some children such as W.M.S. will not receive benefits
because of current state intestacy law. We are hesitant to
strike out ahead of Congress by addressing this increasingly
important issue through a novel interpretation of the Act.
Although Congress can change the Act regardless of what we
decide, we pay greater respect to its legislative primacy by
letting it forge the future path, especially where its previous
22 SCHAFER v. ASTRUE
enactments are at worst ambiguous and most likely prescribe
the approach the SSA is already taking.

Furthermore, upholding the SSA’s interpretation reflects
the basic division of labor between courts and agencies. It is
our duty to interpret and enforce statutes, not to update and
revise them in light of changing technological circumstances.
"Courts are not authorized to rewrite a statute because they
might deem its effects susceptible of improvement."
Badaracco v. Commissioner, 464 U.S. 386, 398 (1984). This
proscription against tinkering applies equally to ambiguous
statutes entrusted by Congress to an agency. Indeed, that is
the whole point of Chevron deference. The SSA has extensive
experience in administering the Act’s survivorship benefits
program, as well as the legal and practical ability to respond
more quickly to changing regulatory circumstances. See Nat’l
Cable & Telecomms. Ass’n v. Brand X Internet Servs., 545
U.S. 967, 981–82 (2005). It also faces political checks from
Congress and the President, likely keeping its policymaking
within the bounds of the democratically acceptable. See Chev-
ron, 467 U.S. at 865–66. In short, this is a case where courts
are well advised to leave the fine-tuning of statutory regimes
to others.

Moreover, this is not a case where an agency seeks to
aggrandize its regulatory powers by aggressively reinterpret-
ing a statute. See, e.g., Brown & Williamson, 529 U.S. at 126,
145–46 (rejecting the FDA’s sudden attempt to regulate
tobacco products). Rather, the SSA seeks only to maintain its
long-held position, one that keeps primary responsibility for
determining child status in states’ hands. This approach
makes perfect sense. Section 416(h)(2)(A) suggests a strong
desire on Congress’s part to use state law in administering the
federal program, as do other provisions in the Act. See 42
U.S.C. § 416(h)(1)(A)(i)–(ii) (generally determining whether
an applicant is the "wife, husband, widow, or widower" of an
insured by reference to state family and intestacy law). The
Act also reflects a desire to work collaboratively with the
 SCHAFER v. ASTRUE 23
states in providing benefits. See, e.g., id. § 421 (allowing
states to request the ability to determine disability status).

Congress’s efforts toward cooperative federalism here are
hardly surprising. Family and inheritance law fall squarely
within the states’ historic competence, see, e.g., Ankenbrandt
v. Richards, 504 U.S. 689, 703–04 (1992), and their input into
the federal benefits program can only redound to its advan-
tage. Moreover, as Schafer’s case perfectly illustrates, these
areas of law are slippery ones, constantly requiring adjust-
ments so as to account for changing technological and social
realities. The states have demonstrated a legislative willing-
ness to account for the new biological world in which we find
ourselves, and we are loath to fault the SSA for maintaining
an interpretation that encourages an adaptive response from
deliberative, democratically-elected bodies in a complex and
evolving area.

Jettisoning the SSA’s approach would "short-circuit what
looks to be a prompt and considered" development of state
law. Dist. Attorney’s Office for the Third Judicial Dist. v.
Osborne, 129 S. Ct. 2308, 2322 (2009). It would also thrust
the federal courts into "a myriad of other issues," id. at 2323,
including policy-driven questions about the child status of
those born to surrogate mothers, those resulting from artificial
insemination of sperm from someone other than the mother’s
husband, and so on, see Capato, 2011 WL 9368, at *5. "The
elected governments of the States are actively confronting"
the challenges posed by modern reproductive technology.
Osborne, 129 S. Ct. at 2322. And "there is no reason to sup-
pose that [federal courts’] answers to these questions would
be any better than those of state courts and legislatures, and
good reason to suspect the opposite." Id. at 2323.

Congress, on the SSA’s view, has created a scheme allow-
ing the full resources of our federal system to be brought to
bear on the novel and complex issues posed by modern repro-
ductive technology. In so doing it has ensured that those chil-
24 SCHAFER v. ASTRUE
dren whom state policymakers decide should stand in the
same shoes as more traditionally conceived children for inher-
itance purposes do so for the analogous purposes underlying
child status under the Act. We will not disturb this long-lived
effort at federal-state cooperation, especially where the SSA’s
construction of the Act seems not only permissible, but also
correct.4

V.

As is often the case when applying the law, our decision is
not without its costs. The tragedies of cancer and heart disease
pulled the Schafer family’s plans out from under them. While
modern medicine allowed Janice Schafer to partially fulfill
some of those plans years later, Virginia intestacy law, as
incorporated by the Act, does render survivorship benefits
unavailable here.

But if sad facts make hard cases, we cannot allow hard
cases to make bad law. Congress instructed the SSA to apply
the Act just as it always has. Its interpretation harmonizes the
Act as a whole, coheres with § 416(h)(2)(A)’s broad text, and
leaves ample work for § 402(d)(1)’s reference to § 416(e). It
comports with Congress’s background understanding of the
Act’s operation and with Congress’s aim of helping those
who unexpectedly lost a parent’s support. And it harnesses the
cooperative effort of state legislatures in confronting the
novel, complex, and contestable issues arising out of new
4
Schafer also contends that the SSA’s interpretation violates the Equal
Protection Clause. Because the statute classifies on the basis of ability to
inherit under state intestacy law, it requires only rational basis review, see
City of Cleburne v. Cleburne Living Ctr., Inc., 473 U.S. 432, 440 (1985),
and it passes muster as rationally related to the government’s interests in
providing benefits primarily to children of deceased wage earners who
relied on those wage earners for support. See Lucas, 427 U.S. at 507–09;
Vernoff v. Astrue, 568 F.3d 1102, 1112 (9th Cir. 2009). Schafer also con-
tends that W.M.S. can inherit under Virginia law, but we find no fault in
the district court’s rejection of that claim.
 SCHAFER v. ASTRUE 25
technology. And even if Congress did not explicitly dictate
the use of state intestacy law, the SSA’s interpretation is cer-
tainly a reasonable one. Chevron deference was made for cir-
cumstances like these, and we cannot invert well-established
administrative law doctrines simply to reach a certain result.5
The judgment of the district court is therefore affirmed.

AFFIRMED

DAVIS, Circuit Judge, dissenting:

The majority contends that "the plain text of th[e]se provi-
sions offers all of the guidance that is needed," Maj. Op. at
11, and I agree. But where the majority believes that "Con-
gress plainly intended" that 42 U.S.C. § 416(h)(2)’s definition
5
For many of these same reasons, we are unpersuaded by our fine col-
league’s dissent. It neglects the fact that (1) Congress has told us quite
clearly how to determine "whether an applicant is the child" in
§ 416(h)(2)(A); (2) that Congress has for years entrusted the SSA with
"full power and authority to make rules and regulations" in carrying out
these very provisions, see 42 U.S.C. § 405(a); (3) that the Supreme Court
has mandated deference to agencies even in ambiguous situations, not to
mention in situations such as this where the Act’s instructions are plain,
see Chevron, 467 U.S. at 843; (4) that Congress sought to harness the
states’ family law experience and updating prowess in an exercise of
cooperative federalism; (5) that the Act’s purpose was to protect those
who had depended on the deceased wage earner, a purpose threatened by
allowing unlimited posthumous access to the insured’s limited total bene-
fits, see 42 U.S.C. § 403(a)(1); (6) that defining "child" as "biological
child" leaves the federal courts with a host of unanswered and unanswer-
able questions stemming from the numerous biological permutations
already possible; (7) that reading § 416(e)(1) as giving child status to all
biological children renders wholly superfluous Congress’s nuanced efforts
at providing child status for out-of-wedlock biological children in
§ 416(h)(3)(C); and (8) that avoiding this problem by reading § 416(e)(1)
as giving child status only to children born in wedlock contradicts that
term’s supposedly plain meaning and nakedly rewrites the Act.
It is unlikely that a reading of the Act that has so many problems would
be the best one. It is unthinkable that it would be the only permissible one
under the Chevron decision.
26 SCHAFER v. ASTRUE
of "child" control, id. at 13, I believe just the opposite: it
could not be more clear that Congress instructed us to apply
42 U.S.C. § 416(e) in this case. And even if § 416(h) were to
apply, proper consideration of § 416(h)(2)(B) would require
that we reverse the district court.

I.

This is not a case in which we must choose between two
competing statutory definitions, for here Congress has chosen
for us. The question before us is whether a posthumously con-
ceived biological child is a "child" within the meaning of 42
U.S.C. § 402(d)(1), the provision guaranteeing benefits to a
deceased wage earner’s children. And § 402(d)(1) itself
begins, "Every child (as defined in section 416(e) of this title)
. . . ." (emphasis added). If that does not make Congress’s
choice clear enough, Congress went on to specifically apply
§ 416(h)(2)’s definition of "child" shortly after, in
§ 402(d)(3): "For purposes of this paragraph, a child deemed
to be a child of a fully or currently insured individual pursu-
ant to section 416(h)(2)(B) or section 416(h)(3) of this title
shall be deemed to be the legitimate child of such individual."
(emphasis added). That is, Congress specifically invoked
§ 416(e) in the provision at issue and went on to specifically
invoke § 416(h) in a neighboring provision; it is difficult to
imagine how Congress could have more clearly indicated that
it understood the difference between the two definitions and
was deliberately choosing to apply § 416(e)’s broader defini-
tion to the right-granting provision.

The majority dismisses this point, claiming "it is not at all
unusual for Congress to refer explicitly only to one section
even though some of that section’s terms are given their full
import by another, unmentioned section." Maj. Op. at 12. It
cites only to 29 U.S.C. § 626(a), granting the Equal Employ-
ment Opportunity Commission investigative authority "in
accordance with the powers and procedures provided in sec-
tions 209 and 211 of this title," because § 209 itself merely
 SCHAFER v. ASTRUE 27
empowers the EEOC to make use of investigatory powers
granted to the Federal Trade Commission by 15 U.S.C. §§ 49-
50. But § 209 specifically cites to the provisions it incorpo-
rates; here, the majority would ignore Congress’s express
directions and insert an implicit citation internal to § 416(e).
It is not clear to me how Congress’s decision not to recopy in
full the powers granted the FTC in title 15 of the Code when
granting these powers to the EEOC in title 42 is in any way
relevant. The question raised by the majority’s approach is
why we should ignore Congress’s explicit instruction that it
was using "child" in §402(d)(1) "as defined in section 416(e)
of this title" and instead look to a different definition in
§ 416(h), especially when Congress specifically cited to
§ 416(h)’s definition in a neighboring provision.

A.

The majority’s approach, while professing deference to the
plain meaning of the statute, does great violence to it. Subsec-
tion 416(e) begins, "The term ‘child’ means (1) the child or
legally adopted child of an individual . . . ." The majority
would demote the first words of § 416(e) from Congressional
instruction to irrelevant "tautology," rendering the first half of
the provision—"the child . . . of an individual"—a nullity.
Maj. Op. at 8-9

Dismissing as tautological a definition of "child" that
includes "the child . . . of an individual" appears to commend
itself to common sense, but it is sensible only if one overlooks
the purpose of statutory definitions. Often they function not
like dictionary entries, which set out the meaning of a word
by deconstructing it into component concepts, listing its con-
notations, or explaining its association to similar words, but
rather as support for a kind of legislative shorthand. Statutory
definitions allow drafters to avoid repeating cumbersome lists
filled with qualifications, as they can instead substitute a sin-
gle word or phrase as an abbreviation. In this case, § 416(e)’s
dense, 521-word description of those who qualify for a sur-
28 SCHAFER v. ASTRUE
viving child’s benefits (including "(2) a stepchild who has
been such stepchild for not less than one year immediately
preceding the day . . . (3) a person who is the grandchild or
stepgrandchild of an individual or his spouse, but only if . . .")
can be abbreviated throughout the subchapter as "child."

If one rejects the majority’s claim of "tautology" as I do
and instead recognizes that Congress chose to include "the
child . . . of an individual" within the list of rights-holders,
one is left with a familiar situation: a court is asked to deter-
mine whether a given factual situation fits within the class
contemplated by the statute. That is, one must determine
whether a posthumously conceived child is "the child . . . of
an individual." Like any word, the word "child" comprises
both a core of relations it clearly encompasses and a hazy
periphery where the label becomes increasingly contested. Cf.
H.L.A. Hart, Positivism and the Separation of Law and Mor-
als, 71 Harv. L. Rev. 593, 606-07 (1958) (famously distin-
guishing "the hard core of standard instances or settled
meaning" from the "penumbra of debatable cases in which
[the term or rule] [is] neither obviously applicable nor obvi-
ously ruled out"). But a biological child birthed by its biologi-
cal mother is clearly at the very core of the meaning of
"child." (Even the majority grudgingly concedes that Con-
gress’s use of the word "child" here "does aim at natural chil-
dren." Maj. Op. at 8.) The fact that appellant’s child was
conceived after his father’s death does nothing to undermine
the fact that, according to ordinary usage, her son is the
"child" of his biological father. For the reasons given by the
Third Circuit in Capato ex rel. B.N.C. v. Comm’r of Soc. Sec.,
___ F.3d ___, 2011 WL 9368 (3d Cir. Jan. 4, 2011), and the
Ninth Circuit in Gillett-Netting v. Barnhart, 371 F.3d 593 (9th
Cir. 2004), I would find that both the plain meaning of the
statute and the relevant legislative history instruct the Com-
missioner to include biological children within the meaning of
"child" as used in § 402(d)(1).
 SCHAFER v. ASTRUE 29
B.

The majority justifies its reading of the statute by empha-
sizing § 416(e)’s "lack of guidance" and asserting that, as that
subsection "cannot provide all of the answers," one must turn
to § 416(h)(2)(A), which "describes precisely how one ought
to determine the meaning of ‘child’ under § 416(e)(1)." Maj.
Op. at 9, 11. The majority’s certainty appears rooted in an
instruction it quotes from § 416(h)(2)(A): "In determining
whether an applicant is the child . . . of a fully or currently
insured individual for purposes of this subchapter, the Com-
missioner of Social Security shall apply such [state] law as
would be applied in determining the devolution of intestate
personal property . . . ." Maj. Op. at 11 . Yet this quotation
is misleading in two respects.

1.

First, it fails to reproduce the second and final sentence of
§ 416(h)(2)(A), which reads, "Applicants who according to
such [state] law would have the same status relative to taking
intestate personal property as a child . . . shall be deemed
such." (emphasis added). That is, § 416(h)(2)(A) requires pre-
cisely the same interpretation of "child" that the majority fled
from in § 416(e)(1). One cannot reasonably compare a claim-
ant’s status under intestacy to the status of "a child" until one
settles on the definition of "child." Thus it makes little sense
to abandon § 416(e)(1) on the ground that the word "child" is
vague in favor of § 416(h)(2)(A), which also requires an
extraneous definition of "child."

Moreover, this second sentence makes it pellucidly clear
that § 416(h)(2)(A) supplements, rather than replaces
§ 416(e)(1). If one finds that a claimant fits within the defini-
tion of the comparator "child," then obviously he or she will
have "the same status relative to taking intestate personal
property as a child." Thus, any "child" within the meaning of
§ 416(e)(1) must, as a straightforward matter of logic, satisfy
30 SCHAFER v. ASTRUE
§ 416(h)(2)(A). That this yields a strange result in a case like
this one, where the relevant state intestacy law treats different
classes of "children" differently and grants property to only
some of them, underscores the difficulties with the majority’s
attempt to cast § 416(h)(2)(A) in the role of deus ex machina.

As a supplementary provision that requires an independent
definition of "child" but provides no additional aid in con-
structing one, § 416(h)(2)(A) is not a helpful alternative to
§ 416(e)(1) for present purposes. Under either provision, one
must still wrestle with the definition of "child."

And it is for this reason that the majority’s attempt to sub-
ordinate § 416(e) to § 416(h)(2)(A) fails. Recognizing that
§ 416(e)’s inclusion of adopted children, stepchildren, grand-
children, and stepgrandchildren must have legal force— the
Commissioner’s regulations include them, see 20 C.F.R.
§§ 403.354, 356-59— the majority suggests that § 416(e)(1)
"clarifies that natural children and legally adopted children
are included," and likewise for sections (e)(2) and (e)(3) and
the other relations. Maj. Op. at 12. But if § 416(h) controls,
exactly where in the text of the provision do these other rela-
tions fit in? As the first sentence only refers to the "applicant,"
it must be in the second sentence, which requires the Com-
missioner to compare the applicant’s would-be status under
intestacy with the status of "a child." And this would undo the
majority’s construction, for it would simply reveal
§ 416(h)(2)(A) to be a supplement, not a replacement, for
§ 416(e): looking to § 416(e) for the definition of "child" in
§ 416(h)(2)(A)’s second sentence, each relation covered in
§ 416(e) would automatically pass, since in each case one
simply asks the redundant question of whether, say, a step-
grandchild inherits in the same manner as a stepgrandchild. If
§ 416(e) overrides the majority’s mistaken view of
§ 416(h)(2)(A) in the exotic case of stepgrandchildren— that
is, if one accepts the Commissioner’s position, as stated at 20
C.F.R. 404.358, and acknowledge that a stepgrandchild is
entitled to benefits regardless of his or her ability to take
 SCHAFER v. ASTRUE 31
under a particular state’s intestacy law— it is not apparent
why the same would not be true in the case of a biological
child.

2.

There is still more to be said. The majority’s use of the
excerpt from § 416(h)(2)(A) is also misleading insofar as it
claims that, given the paragraph’s "shall apply such [state]
law" language, "[t]he text . . . could hardly be more clear"
about which provision applies. Maj. Op. at 9. That is, the
majority appears to read the provision’s mandatory lan-
guage— "for the purposes of this subchapter . . . the Commis-
sioner . . . shall apply such [state intestacy] law"— to mean
that § 416(h)(2)(A) is a requirement that all claimants must
meet. Yet, as the majority concedes earlier in the opinion,
§ 416(h) "also provides three additional gateways to child sta-
tus for those who cannot establish it through § 416(h)(2)(A)’s
intestacy provision." Maj. Op. at 5. Thus § 416(h)(2)(A)’s
"shall apply" language is not to be read as "shall only apply";
rather, Congress is simply instructing the Commissioner to
follow the precise process of applying and comparing situa-
tions under state intestacy law, not restricting the Commis-
sioner to look only at § 416(h)(2)(A).

The majority strains to cast § 416(h)(3), one of these alter-
native paths, as "an exception to a general requirement,"
emphasizing that the provision states that a "son or daughter"
who does not meet the requirements of § 416(h)(2)(A) or (B)
"shall nevertheless be deemed to be the child of such insured
individual." Maj. Op. at 15 (emphasis added by majority opin-
ion). This is apparently meant to shield from attack the major-
ity’s conclusion that Congress was drafting "within the
framework of the background intestacy provisions." Maj. Op.
at 15. The problem is that when Congress repeats this verbal
formula, which first appears in § 416(h)(2)(B), in § 416(h)(3),
paragraph (2)(B) is cited alongside paragraph (2)(A). And
paragraph (2)(B) has nothing to do with state intestacy law:
32 SCHAFER v. ASTRUE
it qualifies the "son or daughter" of parents who "went
through a marriage ceremony resulting in a purported mar-
riage" that, "but for a legal impediment . . . would have been
a valid marriage."

Furthermore, Congress’s decision to label these provisions
as (2)(A) and (2)(B) undermines the majority’s suggestion
that (2)(A) was meant to have pride of place as the "back-
bone" of the definitional framework. Maj. Op. at 14. Rather,
it seems that when Congress instructed that the Commissioner
"shall apply" state intestacy law and then compare the result
to intestate takings of a "child," the mandatory language was
a means of instructing the Commissioner to follow a precise
process, not a way of implicitly undercutting § 416(e). Indeed,
I expect that if Congress were to draft language that gave pri-
macy in defining "child" to a subpart within a subsection enti-
tled "Determination of family status" over the entirety of a
subsection entitled "Children" (within § 416, the "Definitions"
section), such a command would not be left implicit. Cf. Whit-
man v. Am. Trucking Ass’ns, Inc., 531 U.S. 457, 468 (2001)
(Congress does not "hide elephants in mouseholes").

Yet the majority disputes the conclusions that emerge from
the Third and Ninth Circuits’ reviews of legislative history,
which found that § 416(h) was meant to be additive— extend-
ing benefits to the children of unwed parents— rather than an
attempt to supplant and, in places, narrow the scope of bene-
fits promised by § 416(e)’s definition of "child." See Maj. Op.
at 13-15. It quotes from the Senate Finance Committee’s
report on the 1965 amendments to the Social Security Act,
which expanded children’s survivorship benefits (immedi-
ately granting benefits to an estimated twenty thousand peo-
ple) "without regard to whether he has the status of a child
under State inheritance laws." S. Rep. 89-404, pt. 1, at 14
(1965), as reprinted in 1965 U.S.C.C.A.N. 1943, 1958
(emphasis added). This reference, says the majority, demon-
strates that § 416(h)(2)(a)’s intestacy provisions are "the
backbone of all child status determinations." Maj. Op. at 14.
 SCHAFER v. ASTRUE 33
But it is far more likely that Congress was simply referring
back to the intestacy provisions because they were the first to
extend coverage beyond biological children. As the Commis-
sioner has conceded, Congress believed that "all state laws . . .
provided for inheritance of marital children." Br. of Appellee,
at 33; see also 43 Fed. Reg. 57590, 57591 (1998) (finding that
"a ‘child of a valid marriage’ has inheritance rights under the
laws of all states"). The report’s reference to § 416(h)(2)(A)
is simply a reference to an expanded scope of survivorship
benefits reaching beyond marital children; not a suggestion
that paragraph (2)(A) was limiting benefits granted via
§ 416(e)’s "plain language" definition.

In the absence of any indication that the statutory frame-
work is not as it seems, there is no reason to imagine that
§ 416(h) somehow usurps § 416(e), and thus we should not
second-guess Congress when it specifically cites to the defini-
tion of "child" in § 416(e) (and goes on to specifically cite to
the differing definitions in § 416(h) in a following provision).
The majority is correct that "[t]he text . . . could hardly be
more clear," but it misapprehends that meaning. When Con-
gress states in § 402(d)(1), "Every child (as defined in section
416(e) of this title)," that is precisely where one should look
for a definition. (emphasis added).

C.

The majority invokes Chevron deference to the Commis-
sioner’s regulations to bolster its case. See Maj. Op. at 8. Yet
even if the Commissioner can survive Chevron step one, his
interpretation fails at step two because it presupposes that
§ 416(h) supplements, rather than replaces, § 416(e), and thus
his refusal to recognize § 416(e)’s inclusion of biological chil-
dren is not reasonable. The Commissioner, like the majority,
simply misapprehends the statutory language and, most
assuredly, the statutory purpose.

The Commissioner recognizes that § 416(e)’s list of benefi-
ciaries controls; thus he provides in § 404.354, "You may be
34 SCHAFER v. ASTRUE
related to the insured person in one of several ways and be
entitled to benefits as his or her child, i.e., as a natural child,
legally adopted child, stepchild, grandchild, stepgrandchild, or
equitably adopted child." Sections 404.355-39 detail the
means by which one may establish these relationships. Yet
§ 416(h) makes no mention of stepchildren, grandchildren, or
stepgrandchildren; indeed no provision but § 416(e), accepted
as a definition for "child" in §402(d)(1), would authorize the
Commissioner to grant all of these individuals benefits.

As the statute grants the Commissioner rulemaking author-
ity, 42 U.S.C. § 405(a), his regulations are "entitled to defer-
ence if they resolve [any] statutory ambiguity in a reasonable
manner." Coeur Ala., Inc. v. Se. Ala. Conservation Council,
129 S. Ct. 2458, 2469 (2009); see Chevron U.S.A. Inc. v. Nat-
ural Res. Def. Council, Inc., 467 U.S. 837, 842 (1984).
Though I believe that "Congress has directly spoken to the
precise question at issue" and that "the intent of Congress is
clear" on this point, Chevron, 467 U.S. at 842, given that
Congress specifically stated in § 402(d)(1) that it was using
the word "child" "as defined in section 416(e) of th[at] title,"
the Commissioner’s regulations would fail at Chevron step
two. The question is not whether the Commissioner’s results
are reasonable, but whether (assuming statutory ambiguity)
his construction of the statute is reasonable. As it would be
irrational to acknowledge § 416(e)’s force with respect to
every listed relation but one, the Commissioner’s position is
not entitled to deference here.

II.

And even if one did grant Chevron deference, agreeing to
ignore § 416(e) and apply only § 416(h)’s definition of
"child," appellant would be entitled to a reversal. Section
416(h)(2)(B) provides,

[An] applicant shall . . . be deemed to be the child
of [an] insured individual if such insured individual
 SCHAFER v. ASTRUE 35
and the mother or father, as the case may be, of such
applicant went through a marriage ceremony result-
ing in a purported marriage between them which, but
for a legal impediment described in the last sentence
of paragraph (1)(B), would have been a valid mar-
riage.

The last sentence of paragraph (1)(B) specifies that these
"legal impediment[s]" include "the lack of dissolution of a
previous marriage" and "a defect in the procedure followed in
connection with such purported marriage."

It is undisputed that W.M.S.’s parents, Mr. and Mrs.
Schafer, were married in June 1992. If W.M.S. could show
that there was some technical defect in the marriage paper-
work, then, even under the majority’s reading of the statute,
W.M.S. would qualify as a "child" under § 416(h)(2)(B). To
deny him benefits because his parents’ marriage was valid
would be bizarre. This provision again shows that Congress
intended to include children of a valid marriage in its defini-
tion of "child," for what possible purpose would Congress
have for covering children of a technically invalid marriage
but not those of legally valid ones?

To adopt such a reading of the statute would raise serious
constitutional questions. Indeed, appellant has brought an
equal protection challenge to such a construction of the stat-
ute. Though a non-suspect statutory distinction will survive an
equal protection challenge "if there is any reasonably conceiv-
able state of facts that could provide a rational basis for the
classification," FCC v. Beach Commc’ns, Inc., 508 U.S. 307,
313 (1993), such a distinction between valid and technically
invalid marriages might well fail rational basis review.

Here the canon of constitutional avoidance fully applies.
When "choosing between competing plausible interpretations
of a statutory text," courts are to choose in accordance with
"the reasonable presumption that Congress did not intend the
36 SCHAFER v. ASTRUE
alternative which raises serious constitutional doubts." Clark
v. Martinez, 543 U.S. 371, 382 (2005). The simplest choice
here (and, I believe, the legally correct one) would be to give
effect to § 416(e)(1) and find that a biological child is, indeed,
"the child . . . of an individual." But even within the majori-
ty’s view that § 416(h) overrides § 416(e) there exists a plau-
sible reading of § 416(h)(2)(B) that avoids an irrational
distinction favoring technically invalid marriages.

In particular, one might read § 416(h)(2)(B) to implicitly
include valid marriages. Under this view, paragraph (2)(B)
simply extends the benefits granted under paragraph (2)(A)—
a provision thought to include any "child of a valid marriage,"
43 Fed. Reg. 57590, 57591 (1998)— to the children of mar-
riages that Congress believed ought to be considered valid for
its purposes.

Under this interpretation, any child of a marriage (who hap-
pened not to already qualify for benefits under paragraph
(2)(A)) would satisfy § 416(h)’s definition of "child," whether
or not a technicality undermined the marriage’s validity. One
would view the "but for" clause not as listing technical inva-
lidity as a further requirement, but simply as excusing techni-
cal invalidity and extending benefits to reach the children of
such marriages. Manifestly, this was Congress’s intent, and in
fact the paragraph’s logical structure lends itself to such a
reading. As formal logic would have it, it is true of validly
married spouses that they "went through a marriage ceremony
resulting in a purported marriage between them which, but for
a legal impediment . . . , would have been a valid marriage."
42 U.S.C. § 416(h)(2)(B). The "but for" phrase simply
instructs the Commissioner to consider a hypothetical sce-
nario (thus the following subjunctive, "would have been") in
which any technical defects are ignored; that is, in which mar-
riages that fail on account of a technicality are made to look
identical to valid marriages. Though children of valid mar-
riages need not avail themselves of Congress’s liberality,
they, too, would be covered by paragraph (2)(B).
 SCHAFER v. ASTRUE 37
As § 416(h)(2)(B) is drafted in language logically applica-
ble to valid marriages as well as technically invalid ones, and
as the coverage of such was clearly Congress’s intent, this is
at least a plausible reading of the provision. And since this
construction would avoid a serious constitutional question, it
is preferable to the majority’s reading, which would arbitrar-
ily penalize the children of valid marriages because their par-
ents’ marriage proceedings happened not to suffer from some
technical defect.

III.

In this case a husband, facing the prospect of sterility just
four months after marrying, voluntarily left behind his sperm
for a singular purpose: to produce a child with his wife. That
the two of them achieved their joint purpose because she did
exactly what they both contemplated provides scant justifica-
tion to distort statutory meaning, structure, and purpose and
thereby disqualify their offspring from his federal statutory
entitlement. The majority is surely correct in its implied
lament that we live in a "brave new world," one in which the
law lags behind technology, as it ever has. But that truism has
never defined a "hard case." What must be recalled is that
judicial opinions, like the statutes they interpret, are not
merely words arranged on paper. They have real effects on
real people.

Respectfully, I dissent.