Pennsylvania Department of Public Welfare, DAB No. 1278 (1991)

Department of Health and Human Services

DEPARTMENTAL APPEALS BOARD

Appellate Division

SUBJECT:  Pennsylvania Department of Public Welfare                

DATE: October 21, 1991
Docket Nos. 90-38, 90-39 and 90-97
Decision No. 1278

DECISION

The Pennsylvania Department of Public Welfare (State) appealed
determinations by the Administration for Children and Families (ACF)
disallowing federal financial participation (FFP) claimed by the State
under title IV-E of the Social Security Act (Act) for foster care
maintenance payments made by the State. 1/  Docket No. 90-38 involves a
disallowance for fiscal years (FY) 1983 and FY 1984 in the amount of
$3,577,797.  Docket No. 90-39 involves a disallowance for FY 1985 in the
amount of $1,761,404.  Docket No. 90-97 involves a disallowance for FY
1986 in the amount of $2,192,082, which was reduced to $2,107,403.40
based on ACF's reversal of its earlier finding of an erroneous payment
in one case.  ACF brief dated 9/14/90, p. 27.  The disallowances, which
total $7,446,604.40, were based on a random sample of foster care
maintenance payments in one or more counties for each fiscal year.  ACF
found that certain payments did not meet the requirements for title IV-E
funding on one or more grounds, and extrapolated the results of each
sample to all payments made in the counties involved.

On appeal to this Board, the State challenged the statistical sampling
procedure used by ACF.  The State also disputed ACF's findings of
ineligibility with respect to a number of the individual sample
payments.

For the reasons discussed below, we reject the State's challenges to the
statistical sampling procedure.  We also conclude, however, that ACF
incorrectly found erroneous the payments in 35 cases involving children
who were voluntarily removed from their homes for whom there was a
judicial determination prior to October 1, 1978 that removal was
contrary to the child's welfare.  We further conclude that ACF correctly
found the payments erroneous in all of the remaining cases disputed by
the State.  Accordingly, we reverse the disallowances pertaining to the
35 cases but uphold the disallowances to the extent that they are based
on sample payments which were either undisputed or which we find were
erroneous.

Below, we first describe the relevant statutory authority for the title
IV-E program.  We proceed to discuss the State's arguments concerning
the statistical sampling procedure.  We then discuss the State's general
arguments with respect to ACF's findings that certain payments were not
eligible for title IV-E funding because the foster care provider was not
approved or because the child was removed from home pursuant to a
voluntary placement agreement.  Finally, we discuss the remaining
individual payments contested by the State.

Relevant Statutory Authority

Under title IV-E of the Act, FFP is available in foster care maintenance
payments for a child removed from the home of a relative if certain
requirements are satisfied.  This section sets out most of the
requirements at issue in this case.  Other requirements are identified
in the text of the decision where appropriate.

Section 472(a) specifies that foster care maintenance payments shall be
made --

 with respect to a child who would meet the requirements of
 section 406(a) or of section 407 but for his removal from the
 home of a relative (specified in section 406(a)). . . .

Section 406(a) states that --

 [t]he term "dependent child" means a needy child (1) who has
 been deprived of parental support or care by reason of the
 death, continued absence from the home . . . , or physical or
 mental incapacity of a parent, and who is living with his
 father, mother, grandfather, grandmother, brother, sister,
 stepfather, stepmother, stepbrother, stepsister, uncle, aunt,
 first cousin, nephew, or niece, in a place of residence
 maintained by one or more of such relatives as his or their own
 home, and (2) who is (A) under the age of eighteen, or (B) at
 the option of the state, under the age of nineteen and a
 full-time student in a secondary school (or in the equivalent
 level of vocational or technical training), if before he attains
 age nineteen, he may reasonably be expected to complete the
 program of such secondary school (or such training). . . .

Section 407 states that --

 [t]he term "dependent child" shall, notwithstanding section
 406(a), include a needy child who meets the requirements of
 section 406(a)(2), who has been deprived of parental support or
 care by reason of the unemployment . . . of the parent who is
 the principal earner. . . .

Under section 472(a)(1), the child's removal from home must be --

 pursuant to a voluntary placement agreement entered into by the
 child's parent or legal guardian, or . . . the result of a
 judicial determination to the effect that continuation therein
 would be contrary to the welfare of such child and (effective
 October 1, 1983) that reasonable efforts of the type described
 in section 471(a)(15) of this title have been made.

However, children removed from home pursuant to voluntary placement
agreements are eligible only if a state elects to include such children
in its state plan.  See section 472(d).

Section 472(a)(4) further specifies that the child must have --

   (A) received aid under the State plan approved under section
   402 [of title IV-A, the AFDC program] in or for the month in
   which . . . [a voluntary placement] agreement was entered into
   or court proceedings leading to the removal of such child from
   the home were initiated, or (B)(i) would have received such
   aid in or for such month if application had been made
   therefor, or (ii) had been living with a relative specified in
   section 406(a) within six months prior to the month in which
   such agreement was entered into or such proceedings were
   initiated, and would have received such aid in or for such
   month if in such month he had been living with such a relative
   and application therefore had been made.

Section 472(e) prohibits federal funding for foster care maintenance
payments --

 in the case of any child who was removed from his or her home
 pursuant to a voluntary placement agreement . . . and has
 remained in voluntary placement for a period in excess of 180
 days, unless there has been a judicial determination by a court
 of competent jurisdiction (within the first 180 days of such
 placement) to the effect that such placement is in the best
 interests of the child.

Section 472(f) defines the term "voluntary placement agreement" as --

 a written agreement, binding on the parties to the agreement,
 between the State agency, any other agency acting on its behalf,
 and the parents or guardians of a minor child which specifies,
 at a minimum, the legal status of the child and the rights and
 obligations of the parents or guardians, the child, and the
 agency while the child is in placement.

The provisions relating to funding for a child removed from home
pursuant to a voluntary placement agreement were added when the title
IV-A foster care program was transferred to a new title IV-E by the
Adoption Assistance and Child Welfare Act of 1980, Public Law 96-272
(effective October 1, 1980). 2/  Prior to that time, funding was
available only for children removed pursuant to a judicial
determination.  However, Public Law 96-272 also contained a provision,
not codified as part of the Act, which stated that --

 [f]or purposes of section 472 of the Social Security Act, a
 child who was voluntarily removed from the home of a relative
 and who had a judicial determination prior to October 1, 1978 to
 the effect that continuation therein would be contrary to the
 welfare of such child, shall be deemed to have been so removed
 as a result of such judicial determination. . . .

Pub. L. 96-272, section 102(d)(1). 3/

Section 472(b) provides that foster care maintenance payments may be
made only on behalf of a child described in subsection (a) who is --

 (1) in the foster family home of an individual. . ., or (2) in a
 child care institution. . . .

Section 472(c) states that --

 (1) the term "foster family home" means a foster family home for
 children which is licensed by the State in which it is situated
 or has been approved, by the agency of such State having
 responsibility for licensing homes of this type, as meeting the
 standards established for such licensing. . . .


Statistical Sampling Procedure

The State made the following arguments regarding the statistical
sampling procedure used to arrive at the disallowances:

 1.      During FY 1983, ACF's policy did not allow ACF to base a
 disallowance on extrapolation from a sample.

 2.      ACF erred in using a simple random sample of payments
 drawn from Philadelphia and other counties in FY 1983, FY 1984
 and FY 1986 because the errors were not randomly distributed
 among the counties.  (This argument does not relate to FY 1985
 because the only payments reviewed for that year were made by
 Philadelphia County.)  The State contended that ACF should have
 used a stratified sample which treated payments from each county
 separately. 4/

 3.      ACF erred in counting two or more payments which
 pertained to one child as separate errors.

 4.      ACF erred in not using a two-staged approach to
 calculate the disallowance for each fiscal year based on the
 data from the sample.

We address each of these arguments in turn below. 5/

       1. Extrapolation as Basis for Disallowance

The State's argument that the disallowance for FY 1983 was improperly
based on extrapolation from a sample has no merit.  The State cited
Louisiana Dept. of Health and Human Resources, DAB No. 580 (1984), to
support its contention that ACF policy during FY 1983 was to disallow
only individually identified foster care payments.  The State's reliance
on Louisiana is misplaced, however.  In Louisiana, the Board found that,
following unsuccessful efforts to establish, as part of a quality
control system, a tolerance level for errors identified using a
statistical sample of payments made under the Aid to Families with
Dependent Children (AFDC) program, the Office of Human Development
Services (OHDS), of which AYCF was a part, "ended up with a policy
stance which assured states that only individually identified errors
would be held against them."  Id., at p. 2.  The Board therefore
concluded that OHDS could not properly use extrapolation from a sample
to disallow AFDC-FC payments made before a contrary policy was
articulated.  However, as noted previously, Public Law 96-272
established a new title IV-E foster care program and required states to
shift their foster care programs from title IV-A to title IV-E no later
than September 30, 1982 (the end of FY 1982).  Although the title IV-E
foster care program still relies on requirements of title IV-A to
establish eligibility for foster care payments, the effect of Public Law
96-272 was to sever the foster care program from the AFDC program.  See
Louisiana Dept. of Health and Human Services, DAB No. 647 (1985), at p.
8. 6/

Thus, while it is true that, as of the beginning of FY 1983, the policy
for the title IV-A AFDC program was to disallow only for individually
identified payments, this policy was simply inapplicable to payments
made under the completely separate title IV-E. 7/  In the absence of
such a policy, ACF was justified in basing the disallowance of foster
care payments on extrapolation from a sample since, as the Board has
previously stated, "[i]t is well-established in Board and court
precedent that sound statistical sampling methodology can generally be
used to determine the amount of costs properly charged to HHS programs."
Maryland Dept. of Human Resources, DAB No. 1225 (1991), p. 4.  The Board
has upheld disallowances based on extrapolation from a statistical
sample in a variety of HHS programs.

        2. Stratification of Sample

There is no dispute that, as a matter of statistical theory, a
stratified random sample will generally provide more precise estimates
than a simple random sample where stratification is based on a variable
that may affect the outcome of the sample.  See Affidavit of James T.
Massey dated May 9, 1991, p. 2; letter from Arrastia to Garrett dated
9/24/91, p. 5 (citing the Title IV-E Financial Review Guide).
Nevertheless, we conclude that ACF should not be required to use a
stratified random sample to determine the amount of erroneous payments
for the three fiscal years at issue here.  As explained below, ACF
provided unrefuted testimony that use of a stratified random sample
would not result in a reduction of the disallowances in these cases.  We
see no reason to compel ACF to use limited resources to conduct another
title IV-E review when the practical effect of this exercise will be
nil.  Moreover, the State's argument was premised on the assumption that
Philadelphia County had unique problems which resulted in the
identification of substantially more errors for that county than for
other counties.  The State noted specifically that most of the errors
related to voluntary placements arose in Philadelphia.  However, as
discussed later in this decision, we conclude that ACF improperly found
errors in 35 cases involving voluntary placements.  Accordingly, the
factual predicate for the State's argument appears to no longer exist.

Moreover, ACF did not base its disallowance on the "point estimate" (the
single most likely estimate of the true value of erroneous payments) but
rather on the lower limit of a two-sided 90% confidence interval based
on the point estimate. 8/  Thus, by only having to pay the lower limit
of the interval, and not the point estimate, the State is protected with
a 95% degree of confidence from having to pay an amount greater than the
true value of erroneous payments.  ACF alleged and the State did not
dispute that stratification generally tends to narrow the size of
confidence intervals.  Elaborating on this, ACF's consultant stated that
"[a]lthough it is difficult to make judgements about individual cases,
over the long run the stratified sample's confidence intervals will be
smaller. . . ."  Affidavit of James T. Massey, supra, p. 2.  Thus, the
use of a stratified sample here could well narrow the confidence
interval and would consequently be to ACF's, rather than the State's,
advantage.

ACF also argued that "it appears that stratification will have little
effect" in view of the determination of a State consultant, using
combined 1983/1984 audit data, that the the mean errors for Philadelphia
County and for all other counties were "almost identical."  Id.  (The
State's consultant calculated a mean of $614.97 for Philadelphia County
and a mean of $621.41 for all other counties.  Letter from Goldstein to
Manne dated February 28, 1991, p. 2.)  The State did not respond to this
argument.

Furthermore, one of ACF's consultants performed a "poststratification
analysis" on the results of the FY 1983 and 1984 reviews which yielded
larger standard errors for both 1983 and 1984 ($20.88 and $22.34) than
the standard errors generated by the simple random samples on which the
disallowances were actually based ($20.84 and $22.27).  The consultant
concluded that stratification would result in this case in a less
efficient sampling design since the smaller the standard error, the more
efficient the sampling design is.  Affidavit of John A. Gaudiosi dated
5/9/91, pp. 2-3.  The State also failed to respond to this argument.

Finally, we note that, when a criterion for the use of a stratified
random sample advanced by the State is applied, the use of this type of
sample is not justified.  Specifically, the State acknowledged that it
would not have been disadvantaged by use of a simple random sample
unless Philadelphia County was overrepresented in the sample for a
particular fiscal year, i.e., unless the percentage of payments from
Philadelphia in the sample was significantly greater than the percentage
of payments from Philadelphia in the universe.  See TR 99-100.  The
State contended that for FY 1986, 75.46% of the payments in the sample
were from Philadelphia whereas only 60.42% of the payments in the
universe were from Philadelphia. 9/  State's hearing ex. D.  According
to the State, a chi-square test showed that the difference between the
percentage of Philadelphia cases in the sample and the percentage of
such cases in the universe was greater than the expected difference,
i.e., so substantial that a stratified random sample was called for.  TR
84.  (The State did not provide percentages for FY 1983 and FY 1984,
despite its view of the significance of the percentages.)  However, ACF
challenged the State's results, noting that the State had not explained
how the percentages for FY 1986 were obtained.  Using data and methods
which are set out in the record, ACF conducted a chi-square test for all
three fiscal years and concluded that the sample percentages were not
statistically significantly different from what would have been expected
if stratified random sampling had been utilized.  Supplemental Affidavit
of John A. Gaudiosi, dated 9/24/91, p. 2, and Attachment A.  The State
never challenged this conclusion.  Thus, under the State's own theory,
stratification was unnecessary.

Accordingly, we conclude that the State did not show a valid basis for
requiring ACF to use a stratified random sample.

     3. Counting Method

We also disagree with the State's contention that ACF improperly treated
as individual errors two or more payments which involved the same child.
(Since the payments in the sample drawn for each fiscal year were made
on a monthly basis, two or more payments involving the same child could
have been included in the sample.) 10/   The State argued that, under
ACF's approach, the same error (e.g., the lack of a judicial
determination) could be counted twice, and that this could "skew the
data."  TR 38.  The Board has previously addressed this issue and
concluded that the counting method used by ACF was valid.  West Virginia
Dept. of Health and Human Services, DAB No. 1257 (1991).  The Board
there stated in pertinent part:

 The purpose of the . . . review was to estimate the amount of
 foster payments made for ineligible children, not the number of
 court orders which did not satisfy the judicial determination
 requirement.  Thus, ACF properly looked at all of the payments
 in the sample regardless of whether multiple payments pertained
 to a single child or otherwise resulted from a single error.
 This would not bias the estimate, since with a random sample,
 payments resulting from a single error should occur with the
 same frequency in the universe as in the sample.

Id., p. 9.  The Board provided a copy of the West Virginia decision to
the parties and invited them to comment on it.  Letter to parties dated
8/16/91, p. 1.  In the absence of any argument by the State as to why
the rationale for the holding in West Virginia with respect to this
issue was wrong, we reaffirm it here.

      4. Calculation of Disallowance

The State also argued that the method used by ACF to calculate the
disallowance was defective.  We disagree.  ACF first calculated the
average amount of disallowance per sample payment by dividing the total
dollar value of the erroneous payments in the sample by the total number
of payments in the sample.  It then multiplied the resulting figure by
the total number of payments in the universe and took the lower bound of
the confidence interval constructed about that product.  The State
contended that the proper method was to first calculate the average
error for a sample payment by dividing the total dollar value of the
erroneous payments in the sample by the number of erroneous payments in
the sample.  Next, it estimated the percentage of payments in the
universe which were erroneous by dividing the number of erroneous
payments in the sample by the total number of payments in the sample.
It then multiplied together the lower bounds of the confidence interval
constructed about each figure.

ACF cited authority, not controverted by the State, to the effect that
ACF's method was a standard statistical technique.  See affidavit of
James T. Massey dated May 9, 1991, p. 2, citing W.G. Cochran, Sampling
Techniques (3d ed. 1977), section 2.13.  ACF also contended that the
State's suggested method was not statistically valid.  The problem does
not appear to lie with the two-staged approach itself, which yields
precisely the same result as ACF's method if no confidence levels are
computed, but with the construction of a confidence level at each of the
two stages.  The State assumed that by multiplying the two lower
confidence bounds together, it maintained a 90% confidence level, as
required by ACF policy.  However, ACF provided a consultant's opinion
that there was simply no basis for this assumption. 11/

The State did not respond to ACF's consultant's opinion.  The State's
position on this issue was supported only by the consultant originally
retained by the State (see letter from Goldstein to Manne dated 2/28/91)
and not addressed by the State's second consultant, who testified at the
hearing, or in State counsel's briefs.  Accordingly, we conclude that
there is no basis for substituting the method suggested by the State for
the method actually used by ACF.


Voluntary Placement Agreements

ACF found that a number of payments were erroneous solely because the
child in question was removed from home pursuant to a voluntary
placement agreement.  As noted previously, prior to the enactment of
Public Law 96-272, children placed in foster care pursuant to such
agreements were not generally eligible for foster care maintenance
payments.  Moreover, the State did not include children in voluntary
placements in its title IV-E plan during the years reviewed here,
although it had the option of doing so.  ACF ex. A-1.  The State
nevertheless argued that the payments were eligible for title IV-E
funding on the ground that the children should be deemed to have been
removed pursuant to a judicial determination under section 102(d)(1) of
Public Law 96-272.  This provision applies to a child who has been
"voluntarily removed" from home and for whom there has been "a judicial
determination prior to October 1, 1978 to the effect that continuation
therein would be contrary to the welfare of such child. . . ." 12/

The State asserted that in all but one case the requisite judicial
determination was made prior to October 1, 1978, so that the child
should be deemed removed pursuant to a judicial determination and hence
eligible for foster care maintenance payments.  ACF, however, disputed
that a judicial determination was made.  ACF also took the position that
section 102(d)(1) was inapplicable because the agreement pursuant to
which the child was removed did not constitute a voluntary placement
agreement as defined in section 472(f)(2) of the Act.  We address each
of ACF's objections below, concluding that they are unwarranted and that
none of the payments in question are erroneous. 13/

   Contrary to the Welfare Determination

The State contended that the requisite judicial determination was made
when a judge signed a "Placement Agreement" below a pre-printed line
reading "REGISTERED IN THE COUNTY COURT OF PHILADELPHIA JUVENILE
DIVISION."  The record establishes that, after a placement agreement had
been signed by the child's parent (or parents) and the social worker, it
would be forwarded to the Philadelphia Family Court with a motion asking
for approval of the placement. 14/  The court would log in the agreement
and place it on a motions list.  Once or twice a month, depending on the
number of cases, a representative of the Department of Public Welfare
would appear in open court and present the cases on the motions list to
the judge.  The representative would briefly state the reason for the
child's placement, after which the judge would sign the placement
agreement as indicated  above.  The judge did not necessarily sign every
placement agreement presented to the court, however.  State's ex. 6; TR
15-17.  The State submitted a letter from Judge Nicholas A. Cipriani,
who signed many of the placement agreements in question, which stated in
pertinent part that "[a]t the court hearing, the judge would sign the
voluntary placement agreement finding that the foster care placement was
in the best interests of the child."  State ex. 6, 1st page.  Judge
Cipriani also testified at the hearing in these appeals that when he
reviewed the placement agreements, he was "determining what was in the
interest of the child, whether or not the child should be in placement."
TR 17. 15/

ACF responded that there was no evidence on the face of the placement
agreements that the court's signature constituted a judicial
determination to the effect that continued residence in the home of a
relative would be contrary to the child's welfare.  In ACF's view, these
documents showed only that the State "registered" the placement
agreements with the court.  ACF thus argued that the State could not use
the statements of judicial officials such as Judge Cipriani "to prove
that the court took an action not reflected in the record."  ACF brief
dated 9/14/90, p. 19.

Contrary to ACF's view, however, we conclude that the State properly
established that the court's signature on the placement agreement
constituted a judicial determination.  When the meaning of a document is
unclear on its face, extrinsic evidence is properly considered to show
what was intended.  Here, the significance of the court's signature is
unclear.  To say that the signature indicates only that the State
"registered" the placement agreement with the court is not an answer
since the meaning of registration is itself unclear.  Thus, we see no
reason why the testimony of Judge Cipriani should not be used to clarify
this ambiguity.  ACF did not suggest any reason why we should not find
his testimony credible.  Judge Cipriani held the office of Judge of the
Court of Common Pleas for over 20 years and served as Chairman of the
Pennsylvania Supreme Court Rules Committee, which was responsible for
adopting rules of civil procedure for domestic relations matters.  TR
14.  Moreover, as previously indicated, he signed many of the placement
agreements in question.  He was thus well-qualified to speak to the
matters at issue.

As indicated previously, Judge Cipriani's testimony (which is
corroborated by written statements by him and others at State ex. 6)
establishes that each placement agreement in question came before the
court on a motion by the State for approval of the placement agreement
and that, after an oral proceeding at which the reasons for placement
were presented by the State, the judge signed the agreement if he was
satisfied that the placement was in the "best interest" of the child.
As a matter of policy, ACF treats a determination that placement is in
the best interests of the child as satisfying the requirement for a
contrary to the welfare determination.  See ACYF-PIQ-86-02, dated May 8,
1986, pp. 5-6.  Moreover, while Judge Cipriani was not the only judge
who signed the placement agreements, he testified that the procedure
pursuant to which the agreements were signed was followed by the two
other judges involved.  TR 28.  Accordingly, we conclude that the
judge's signature reflected the judicial determination required by
section 102(d)(1) of Public Law 96-272 in each of the cases at issue
here.

This case is distinguishable from the situation presented in
Pennsylvania Dept. of Public Welfare, DAB No. 1181 (1990), where the
Board found certain court orders inadequate on the ground that they did
not contain an explicit contrary to the welfare determination.  Although
an explicit determination was also lacking here, it is clear from the
circumstances under which the placement agreement was signed by the
judge that the signature had a particular meaning: that in the judge's
view continuation in the home was contrary to the child's welfare.
Moreover, while the court orders described in DAB No. 1181 served to
transfer custody of the child even absent a contrary to the welfare
determination, there is no evidence that the signing by the court of the
placement agreements in question here served any function other than
that asserted by the State.

This case is also distinguishable from Nebraska Dept. of Social
Services, DAB No. 1250 (1991), and West Virginia Dept. of Health and
Human Services, DAB No. 1257 (1991), in which the Board found that the
requisite judicial determinations (contrary to the welfare
determinations or reasonable efforts determinations or both) were not
made in a number of individual sample cases where the original court
orders did not refer to the required determinations.  In each of these
decisions, the Board found that the evidence was insufficient to
establish that the requisite determination was made at the time of
removal.  Specifically, the states did not establish that nunc pro tunc
orders issued years later were issued to correct a mistake in the
original orders rather than to add findings not previously made.
Moreover, the original court orders in Nebraska and West Virginia
clearly lacked any reference to the required determinations even though
a court order would ordinarily reflect any material findings made by the
court.  In the appeals now before us, the issue is the meaning of the
court's signature on voluntary placement agreements signed during a
period when there was some question as to whether any children in
voluntary placements were eligible.  In circumstances where a statute is
amended retroactively to cover children not clearly covered previously,
the State still has a burden to document the allowability of its costs
but reasonably should not be held to the same strict standards about the
nature of the evidence it must present. 16/

ACF nevertheless argued that there was no judicial determination here
because neither the parents nor a child advocate were present at the
court proceeding which resulted in the court's signing the placement
agreement.  ACF contended that since the judge thus relied solely "on
whatever information the Department of Public Welfare . . . chose to
give him, . . . [the judge] would have been unable to reach a considered
judgment whether continuation in the home would have been contrary to
the child's welfare."  Letter from Arrastia to Garrett dated 9/24/91, p.
2.

We disagree.  The statutory requirement is for a "judicial
determination," which is nothing more than a determination made by a
judge acting in his official capacity.  The only restriction in the
statute on how that determination must be made is in section 475(5)(C),
which states that "procedural safeguards shall also be applied with
respect to parental rights pertaining to the removal of the child from
the home of his parents. . . ." However, section 475(5)(C) was enacted
as part of Public Law 96-272, effective October 1, 1980, and was not yet
in existence when the judicial determinations in question here were
made.  Thus, even assuming that section 475(5)(C) requires the presence
at a hearing of parents who have signed a voluntary placement agreement,
we conclude that it would be unfair under the circumstances here to find
a judicial determination inadequate on the ground that the parents were
absent. 17/

Our conclusion that the judicial determination required by section
102(d)(1) was made in the cases in question here is further supported by
the legislative history of that section.  In introducing it as an
amendment on the floor of the Senate, Senator Moynihan characterized the
cases to which the amendment would apply as "cases where there is no
question that the court had approved the child's foster care status. . .
."  125 CONG. REC. 29545 (1979).  Moreover, Senator Javits stated that
"[o]ur amendment would clarify section 408 of the Social Security Act by
providing that once a court has ratified the appropriateness of the
placement of an AFDC child in foster care, payments for that child's
care would be eligible for Federal reimbursement."  Id.  There is no
indication in either statement that any particular form of court
proceeding was required in order for the court to approve or ratify the
child's placement.

       Voluntary Placement Agreement

As indicated above, ACF also argued that section 102(d)(1) of Public Law
96-272 did not apply here because the agreements pursuant to which the
children in question were removed from home did not constitute voluntary
placement agreements within the meaning of the Act. 18/  This argument
is flawed, however, because it assumes that the Act's definition of
"voluntary placement agreement" applies here.  This definition is not
directly applicable since section 102(d)(1) refers to a child "who was
voluntarily removed from the home. . . ," and not specifically to a
voluntary placement agreement.  Moreover, since section 102(d)(1)
requires that a contrary to the welfare determination be made prior to
October 1, 1978, the child must have been removed from home prior to
that date.  The section 472(f)(2) definition of "voluntary placement
agreement" did not become law until October 1, 1980 (the effective date
of Public Law 96-272).  ACF would thus have us apply this definition to
actions which took place before it became law.  This is particularly
unreasonable in view of the fact that the definition goes beyond a basic
definition of the term to require that an agreement specify "the legal
status of the child and the rights and obligations of the parents or
guardians, the child, and the agency while the child is in placement."
Although the placement agreements in question here were not this
detailed, they clearly indicate that the children were removed from home
on a voluntary basis.  Accordingly, we conclude that section 102(d)(1)
applies and that the payments made on behalf of these children were not
erroneous.


Approved Foster Homes

ACF determined that payments in a number of FY 1983, 1984, 1985 and 1986
cases were erroneous because the home in which the child was placed was
not licensed by the State as a foster family home or approved by the
State as meeting its licensing standards for foster family homes.  The
State contested ACF's findings in a number of cases, submitting
documentation which it claimed showed that the homes in question had in
fact been approved.  We address the adequacy of some of this
documentation in our discussion of the individual cases at the end of
this decision.  However, we address here the adequacy of the
documentation consisting of the affidavits of social workers or social
work supervisors who signed the forms authorizing payment in the cases
in question. 19/  Each affidavit, which is accompanied by a Form 85.18
or Form 85.29 (a later version of Form 85.18), states that the affiant
would not have signed the form if the home had not been an approved
home.  (The form itself includes the name and address of the foster
parent, the date of placement, the rate of payment, and in some cases,
comments concerning the reason for placement.)  The State also presented
testimony by a former social work supervisor for the Philadelphia
Department of Human Services (which handled the cases in question) to
the effect that before payment was authorized in a foster care case, the
social worker was required to make a home visit to evaluate the home.
The witness testified that the existence of the Form 85.18 thus showed
that the home had been approved.  TR 118.

We do not agree that a signed Form 85.18 or Form 85.29 establishes that
the foster home was approved.  As the witness herself acknowledged, the
conclusion that a home was approved because payment was authorized
depends on the assumption that the departmental policy requiring a home
to be approved before issuance of a payment was actually followed in a
particular instance.  Id.  Since the issue here is whether a home for
which payment was made was approved, the State essentially asks the
Board to assume the truth of the fact to be proved, which we cannot
reasonably do.  Accordingly, we conclude that the payments in those
cases for which this was the only documentation submitted were
erroneous. 20/


Individual Cases

The State contested ACF's findings with respect to a total of 63 cases.
21/  The 36 cases which involve solely the issue of voluntary placements
are discussed earlier in this decision.  As indicated there, we find
that title IV-E funds were properly used for the payments in 35 of those
cases.  Also discussed earlier are 12 cases involving approved homes in
which the State submitted certain documentation which we find
inadequate.  We discuss below the remaining 15 cases, all of which we
find involved erroneous payments.

Sample Case #190 (FY 1983)

ACF found the payment in this case erroneous on the ground that the
provider was not eligible.  ACF cited, as authority for its finding, the
definition of "foster family home" in section 472(c) of the Act.  The
State contended that the home was approved, and submitted in support of
its contention a 2/25/82 letter from a social worker, addressed "To Whom
It May Concern," which stated that the individual in question "is the
approved caretaker" of the child on whose behalf the payment was made.
State ex. 9.

We find that this letter is not adequate to establish that the provider
was approved.  While the letter states that the provider is the
"approved caretaker," it does not specifically indicate that the
provider was approved as meeting licensing standards by the State agency
having responsibility for licensing foster family homes.  Moreover, the
letter does not indicate the period for which the provider was approved.
Accordingly, we affirm ACF's finding of an erroneous payment.

Sample Case #240 (FY 1983)

ACF found the payment in this case erroneous on the ground that there
was no deprivation of parental support, citing sections 406(a) and
407(a) of the Act.  Under these provisions, deprivation of parental
support can be established by the death, continued absence from the
home, or physical or mental incapacity of a parent, or the unemployment
of the parent who is the principal earner.    ACF asserted that "the
case file revealed that at the time of the child's removal, both parents
were living together, employed and were not mentally handicapped."  ACF
brief dated 9/14/90, p. 21.  ACF later indicated that it was relying on
its reviewer's notes, as the case file itself was in the State's
possession.  ACF brief dated 11/8/90, p. 9.

In support of its position that deprivation existed, the State submitted
a copy of a court order dated 12/2/81 which terminated the parental
rights of the child's mother and father and awarded custody to the
Allegheny County Institution District.  State ex. 10.  The State argued
the child was deprived because the father no longer had a legal duty to
support her.  TR 121-122.  The State also submitted an eligibility
redetermination form for the child which appears to be dated 3/31/84 and
which has a box checked to indicate that the child was receiving AFDC at
the time the court action removing the child was initiated (on 4/26/72).
State's ex. 30.  The State argued that this form should be accepted as
evidence that the child was deprived since the State was not required to
retain eligibility records dating back to 1972.  Letter from Manne to
Garrett dated 9/13/91, p. 3.   We conclude that the State has not
established that the child was deprived within the meaning of the
statute.  The State's reliance on the termination of parental rights is
misplaced.  Since the termination occurred after the court action
removing the child from home, this document does not speak to the
question of whether the child was deprived at the time of removal.  The
State's reliance on the eligibility redetermination form is also
misplaced under the circumstances here.  This form arguably establishes
a presumption that deprivation existed since eligibility for AFDC is
conditioned on deprivation.  However, the presumption is not conclusive
since AFDC payments, like title IV-E payments, can be erroneous.  Here,
ACF specifically found that the case file ruled out the existence of any
deprivation factor -- both parents were living together, were employed,
and were not mentally handicapped.  The State, although it had
possession of the case file, did not offer any basis for challenging
this finding in response to the Board's specific inquiry regarding the
matter. 22/  A presumption that deprivation existed was thus not
justified.  Accordingly, we conclude that the payment in question was
erroneous. 23/

Sample Payment #304 (FY 1984)

ACF found the payment in this case erroneous on the ground that the
child was not removed from the home of a specified relative, citing
sections 472(a)(1) and 406(a) of the Act.  The State asserted in
response, and ACF did not dispute, that the child was originally living
with his mother in the child's paternal grandparents' house and that the
child was then placed with his mother in a foster home.  An excerpt from
the caseworker's notes provided by the State indicates that there was a
court order directing the child's as well as the mother's placement.
State ex. 12.

However, ACF maintained that the "infant" in a "mother/infant placement"
was not eligible for federal funding under title IV-E.  ACF asserted
that the Act was amended effective April 1, 1988 to permit a state to
make foster care maintenance payments on behalf of infants in
mother/infant placements and argued that it was clear from the
legislative history of the amendment that such payments were not
previously permitted.  TR 126-127. 24/  The State contended, however,
that the amendment applied to the situation "where only the mother has
been placed in foster care and the child lives in foster care as
incident to the mother living there" and that title IV-E funding has
always been available for infants who are in the legal custody of the
foster parent.  Letter from Manne to Garrett dated 10/13/91, p. 2.

We find that the legislative history relied on by ACF supports its view
that the "infant" in a mother/infant placement was not eligible for
foster care maintenance payments during the period in question here.
The report on the House bill (which was ultimately enacted with only a
changed effective date) described "Present law" in relevant part as
follows:

 In some teenage foster care cases, the foster child is herself
 the parent of an infant child who lives in the same foster
 family home or a specialized residential program for teen
 mothers and their infants.  The infant child of a foster child
 living in the same foster care home or institution as the
 infant's mother, is not eligible for Federally assisted foster
 care.  This is because the infant technically has not been
 removed from the home of the parent and placed in foster care.
 Such children of foster care children may be eligible as a
 child-only AFDC unit but at a payment level much less than the
 cost of specialized foster care programs for foster care
 teenagers and their infants.

H. R. Rep. No. 319 (II), 100th Cong., 1st Sess. 909 (1987).  This
unequivocally states that a child who was in foster care with his mother
was not eligible for foster care maintenance payments.  Moreover, this
statement is not limited to infants who are in foster care only incident
to their mother's placement, as asserted by the State.  Accordingly, we
conclude that the payment in question was erroneous.

Sample Case #360 (FY 1984)

ACF initially found the payment in this case erroneous on the ground
that the court order did not indicate that reasonable efforts were made
to prevent the child's removal from home, as required by section
427(a)(1) of the Act.  The payment in question was made in April 1984,
during the child's original foster care placement.  The child was
returned to his mother on November 16, 1984, but was later removed
pursuant to a court order dated December 13, 1984 which approved the
"Recommendation of the Master" appointed in the case.  The "Master's
Findings of Fact and Recommendation" included a finding that
"[r]easonable efforts have not been successful to avoid further
placement."  State ex. 13.  The State contended that the December 13,
1984 order satisfied the requirement for a reasonable efforts
determination.  However, ACF responded that title IV-E funding is
available only if the reasonable efforts determination is made before
the date of the payment.

We find that the requirement for a reasonable efforts determination was
not met here.  Section 472(a) requires that the child's removal from
home be "the result of a judicial determination to the effect . . . that
reasonable efforts of type described in section 471(a)(15) have been
made. . . ."  (Emphasis added.)  In this case, only the child's second
placement resulted from the December 13, 1984 determination that
reasonable efforts to avoid further placement had been unsuccessful.
The Act requires, however, that the child's original foster care
placement, during which the payment in question was made, have resulted
from a reasonable efforts determination.  Accordingly, we conclude that
this payment was erroneous.

Sample Case #455 (FY 1984)

ACF found the payment in this case erroneous on the ground that either a
social security number was missing or application for a social security
number was made after the date of care, citing section 402(a)(25) of the
Act.  That section requires that all members of an assistance unit have
a social security number or have applied for one as a condition of
eligibility for AFDC benefits under title IV-A.  Section 472(a)(4) of
the Act provides that a state may make foster care maintenance payments
only with respect to a child who was either receiving or eligible to
receive title IV-A funds.  The State did not dispute that the
requirement for a social security number was applicable here, but
contended that the requirement was a technical one and that the State
cannot be sanctioned for an error which was based on the failure to
comply with this requirement.

We find no basis for concluding that the requirement for a social
security number should not be enforced by sanctioning the State.  The
Act clearly makes compliance with this requirement a condition of
eligibility for title IV-E.  See Maryland Dept. of Human Resources, DAB
No. 1225 (1991), p. 10.  Accordingly, we conclude that the lack of a
social security number for the child or a member of the child's
assistance unit was a proper basis for finding the payment in question
erroneous.

Sample Case #5094 (FY 1985)

ACF found the payment in this case erroneous on the ground that the
child was over age 18 and not a full-time student, citing section
406(a)(2) of the Act.  The State submitted a letter dated 6/6/85
forwarding to the State agency a copy of the child's "final grades from
the Philadelphia Community College."  State ex. 15.  This indicates that
the child was attending community college during FY 1985.  However, a
community college is a post-secondary school, not a secondary school, as
required by section 406(a)(2).  Accordingly, we conclude that the
payment in question was erroneous.

Sample Case #5241 (FY 1985)

ACF found the payment in this case erroneous on the ground that the
provider was not eligible, citing the definition of "foster family home"
in section 472(c) of the Act.  The State contested this finding and
submitted page 3 of a document entitled "Annual Performance Evaluation."
State ex. 17.  However, we fail to see how this document establishes the
provider's eligibility.  The document contains instructions to check one
of three boxes "to indicate home has been approved/certified,
provisionally approved/certified, or not approved/not certified."  None
of the boxes are checked.  Accordingly, we conclude that the payment in
question was erroneous.

Sample Case #5249 (FY 1985)

ACF found the payment in this case erroneous on a number of grounds
which it later conceded were not justified.  ACF brief dated 9/14/90, p.
24; TR 128.  However, ACF contended that the payment was still erroneous
because the child in question was placed in foster care pursuant to a
voluntary placement agreement.  ACF brief dated 9/14/90, p. 24.  In
response to the Board's inquiry whether the State could "provide any
documentation regarding this case which refutes the Agency's contention
. . . that the child was removed from home under a voluntary placement
agreement," the State stated that it "has no additional documentation on
this case and therefore concedes it is an error."  See letter from Board
to parties dated 8/16/91, enclosed question 7; letter from Manne to
Garrett dated 9/13/91, p. 3. 25/  Notwithstanding the State's apparent
concession of an error based simply on removal pursuant to a voluntary
placement agreement, there would be no error if the State could show
that the case fell within the exception in section 102(d)(1) of Public
Law 96-272.  However, in the absence of any documentation showing that a
judicial determination to the effect that removal from home was contrary
to the child's welfare was made prior to October 1, 1978, we conclude
that the payment in this case was erroneous.

We note that the State argued initially that a finding of an erroneous
payment could not be based on the fact that the case involved a
voluntary placement because ACF was precluded from raising new grounds
not stated in the disallowance determination.  This argument has no
merit.  The Board has previously held that a party may raise new
arguments before the Board as long as the other party has an adequate
opportunity to respond to those arguments.  See New York State Dept. of
Social Services, DAB No. 390 (1983), p. 6.  Here, the State clearly had
ample opportunity to address the voluntary placement issue generally,
since other cases were initially disallowed on the ground that they
involved voluntary placements.  Moreover, ACF raised the voluntary
placement issue in this case in its response to the State's initial
brief.  Thus, the State could have provided additional information and
argument specifically with respect to this case in its reply brief, at
the hearing, and in its post-hearing submission.

Sample Case #5269 (FY 1985)

ACF found the payment in this case erroneous on the ground that the
provider was not eligible, citing the definition of "foster family home"
in section 472(c) of the Act.  The State took the position that the case
should not have been included in the sample because the child in
question moved out of Philadelphia in the middle of the month for which
payment was made.  TR 122-123.  The FY 1985 review covered only payments
made by Philadelphia County.

We are not persuaded that the change in the child's place of residence
was a basis for removing the payment from the sample.  The record
indicates that the child moved from a foster home in Philadelphia to a
foster home in Pittsburgh in the summer of 1984 (before the payment in
question was made).  However, the record further indicates that the
Philadelphia Department of Human Services planned to request that the
social services agency in Pittsburgh provide "courtesy supervision" of
the child's case but otherwise to retain responsibility for the case.
State ex. 19.  Thus, it appears that the entire payment was made by the
Philadelphia Department of Human Services, so that the payment in
question was properly included in the sample.  Since the State did not
dispute ACF's finding that the provider was not eligible, we affirm that
finding and conclude that this payment was erroneous.

Sample Case #5270 (FY 1985)

ACF found the payment in this case erroneous on the ground that the
provider was not eligible, citing the definition of "foster family home"
in section 472(c) of the Act.  The State challenged ACF's finding and
submitted the "face sheet" of a document entitled "Application for Home
Study."  State ex. 20.  However, we fail to see how this document
establishes the provider's eligibility.  The fact that a provider
applied for a home study does not necessarily mean that a home study was
conducted or that the provider was approved.  Even if the document was
probative of approval, the document, which is undated, would not
establish that the provider was approved during the fiscal year in
question.  Accordingly, we conclude that the payment in question was
erroneous.

Sample Case #5284 (FY 1985)

ACF found the payment in this case erroneous on the ground that the
provider was not eligible, citing the definition of "foster family home"
in section 472(c) of the Act.  The State disputed this finding and
submitted page 3 of a document entitled "Annual Performance Evaluation."
State ex. 21.  A box is checked on the document indicating that "[t]his
foster home has been provisionally approved/certified."  However, there
is insufficient information to establish that the home was approved,
provisionally or otherwise, when the payment in question was made.  ACF
asserted, and the State did not dispute, that the payment was for
October 1984.  Agency's brief dated 9/14/90, p. 25.  While the Annual
Performance Evaluation was signed by the social worker and the foster
parent on 9/21/83, there is no evidence in the record as to how long the
provisional approval lasted or whether the foster home ever came into
compliance.  Accordingly, we conclude that the payment in question was
erroneous.     Sample Case #60 (FY 1986)

ACF found that the payment in this case was erroneous on the grounds
that the child was removed from home pursuant to a voluntary placement
agreement and that the child was over 18 and not a full-time student,
citing section 406(a)(2) of the Act with respect to the latter grounds.
Although the State argued generally that the cases involving voluntary
placements were eligible for title IV-E funding because the judge's
signature on the placement agreement constituted a judicial
determination, the State did not provide a copy of the placement
agreement in this case.  Accordingly, we conclude that ACF properly
found the payment in this case to be erroneous on this ground alone.

We also conclude that the other grounds for ACF's finding were proper.
The State took the position that, although the child in question here
had graduated from high school, he was enrolled "in the equivalent level
of vocational . . . training" at the McCarrie School of Health Sciences
and Technology within the meaning of section 406(a)(2).  Letter from
Manne to Garrett dated 9/13/91, p. 4.  However, even assuming that the
McCarrie School provided an equivalent level of vocational training, the
State itself admitted that the documentation which it submitted showed
that the child was not enrolled in the McCarrie School until September
1986, after the fiscal year in question here.  State brief dated
10/18/90, p. 6.  Moreover, the documentation indicates that the child
was 19 at the time the child enrolled in the McCarrie School, and thus
could not "reasonably be expected to complete the program of such . . .
training" "before he attains age nineteen," as required by section
406(a)(2).  Accordingly, we conclude that the payment in question was
erroneous on these grounds as well.

Sample Case #115 (FY 1986)

ACF initially found the payment in this case erroneous on the ground
that a court order was not initiated within six months of the child's
removal from home, in accordance with section 472(a)(4) of the Act.  ACF
later asserted in addition that the payment was erroneous because the
case involved a mother/infant placement.  ACF brief dated 9/14/90, pp.
26-27.  The State responded that it had no documentation to address the
latter issue, which it argued ACF was precluded from raising.  We find
no merit in this argument since the State had ample opportunity to
respond to the new grounds advanced by ACF.  (See discussion of Sample
Case #5249 above.)   Accordingly, based on our discussion of
mother/infant placements with respect to Sample Case #304, we find that
the payment in question was erroneous on this ground alone. 26/

Sample Case #242 (FY 1986)

ACF found the payment in this case erroneous on the ground that the
child was not removed from the home of a specified relative, as required
by section 472(a).  The State contested this finding on the ground that
the child was removed from the care of his mother.  In support of its
position, the State provided a form captioned "Information for Court
Hearing" which indicates that the child was committed by the court to
the Department of Human Services retroactive to September 18, 1985, with
a review of the case scheduled for June 1986.  The form further
indicates that the child was "with his mother in mother-and-baby foster
care. . ." before the court hearing (on December 26, 1985) and that this
placement was not changed by the court.  State ex. 26.  ACF argued,
however, that title IV-E funds were not available for mother/infant
placements during the relevant time period.  The State did not dispute
that the child was in this type of placement when the payment was made.

As we stated in our discussion of Sample Case #304, we agree with ACF
that the "infant" in a mother/infant placement was not eligible for
foster care maintenance payments prior to April 1, 1988.  Accordingly,
we conclude that the payment in question was erroneous.

Sample Case #267 (FY 1986)

ACF found the payment in this case erroneous on the ground that the
child was over age 18 and not a full-time student, citing section
406(a)(2) of the Act.   The State contested this finding and submitted
notes from the child's case record written by a social work supervisor
which state that the child "attends Richmond Technical Community College
in Hamlet, N.C. for pre-nursing studies. . . ."  State ex. 27, p. 1.
The notes also establish that the time in question was the school
quarter ending in December 1986.  Id., p. 2.  ACF asserted, and the
State did not dispute, that the payment was made in September 1986.  ACF
brief dated 9/14/90, p. 28.  Since the documentation submitted by the
State indicates that the child was attending a community college at that
time, she was not a full-time student in a secondary school within the
meaning of section 406(a)(2).  Accordingly, we conclude that the payment
in question was erroneous.

Conclusion

For the foregoing reasons, we reverse the disallowances to the extent
that they pertain to the 35 cases for which the State provided placement
agreements signed by a judge prior to October 1, 1978, and uphold the
remaining amount of the disallowances.

 

 _____________________________ Judith A. Ballard

 

 _____________________________ Norval D. (John) Settle

 

 _____________________________ Donald F. Garrett Presiding Board
 Member


1.  This disallowance was originally issued by the Administration for
Children, Youth and Families (ACYF).  Effective April 15, 1991, ACYF was
one of several agencies combined into the Administration for Children
and Families.

2.  The foster care program under title IV-A was repealed effective for
any state at the time a state plan under title IV-E became effective,
but no later than September 30, 1982.  Pub. L. 96-272, section
101(a)(2).

3.  It appears from the legislative history that the October 1, 1978
date in section 102(d)(1) was used because it was the date when New York
State became aware that ACF interpreted section 408 of the Act as
precluding federal foster care payments for children initially placed in
foster care without a court order even if a court later approved the
placement.  The purpose of the amendment was to permit New York to
retain federal funding for payments it made to such children at a time
when it "in perfectly good faith, assumed the claims it filed were
legal."  125 CONG. REC. 29545 (1979) (statement of Senator Moynihan).
However, the legislative history further indicates that Congress
intended similarly situated states to benefit from the amendment.  Id.
As discussed later, we find that certain payments made by Pennsylvania
fall within the ambit of section 102(d)(1).

4.  ACF characterized the type of stratified random sample argued for by
the State as a proportional stratified random sample.

5.  The State withdrew its contention that it was necessary to correct
for a skewed distribution.  See transcript of 8/28/91 hearing (TR), 93,
111.

6.  ACF pointed out that the State submitted a separate state plan for
its title IV-E program in 1982 and in January 1983 amended its cost
allocation plan to provide for recovery of the costs of administering
the title IV-E program.  ACYF brief dated 9/14/90, p. 31.  This
indicated the State's awareness that the title IV-E foster care program
was no longer part of title IV-A.

7.  Since they are made under a separate program, foster care payments
are not a part of the AFDC quality control system, and the rationale for
not permitting extrapolation as a basis for AFDC disallowances during
the time in question does not apply.

8.  A confidence interval is a range of amounts (on both sides of the
point estimate) which can be said, by a percentage of confidence, to
include the true value being measured.  For example, in a two-sided 90%
confidence interval, there is 95% confidence that the true value would
be higher than the lower limit of the interval.  See, e.g., Affidavit of
James T. Massey, supra, para. 7, p. 2.

9.  Although the percentages of cases in the sample from the other two
counties reviewed during FY 1986 did not correspond to the percentages
of cases in the universe from those counties, the State did not contend
that this required ACF to have separate strata for all counties
reviewed.  See TR 61, 101-102; see also letter from Goldstein to Manne
dated 2/28/91, p. 1.

10.  We assume for purposes of this discussion that there were some
instances in which this actually occurred, although this cannot be
determined from the record before us.

11.  ACF's consultant stated specifically that--

 [i]n order to construct confidence intervals about the product
 of the average positive disallowance times the estimated number
 of positive disallowances in the population, one must determine
 the sampling distribution of the product.

Affidavit of James T. Massey dated May 9, 1991, p. 2;  see also
affidavit of John A. Gaudiosi dated May 9, 1991, p. 3.

12.  The State's title IV-E plan effective October 1, 1983 incorporated
this provision with an additional requirement, stating that such a child
--

 shall be deemed to have been so removed as a result of such
 judicial determination if and from the date that a case plan and
 review have been made determining the child's need of foster
 care . . . .

ACF ex. A-1, 9th page (syntax garbled in the original).  ACF did not
contend that this additional requirement was not satisfied here.

13.  The State originally contested 36 cases on this basis and submitted
copies of placement agreements for each case.  See State ex. 8 (FY 1983
cases), ex. 11 (FY 1984 cases) and ex. 23 (FY 1986 case).  The case
later conceded by the State appears to be Sample Case #150 (FY 1983), in
which the judge's signature on the placement agreement is dated
10/26/78.  State ex. 8.  This case would not qualify for title IV-E
funding under section 102(d)(1) because the judicial determination was
not made prior to October 1, 1978.  Moreover, we see no other basis on
which title IV-E eligibility can be found.  Under section 102(d)(1),
title IV-E funding is available only under limited conditions for
children removed from home pursuant to voluntary placement agreements
before such funding was authorized by law and included in a state's
title IV-E plan.  This case fails to satisfy those conditions.

14.  The State did not provide copies of any of the motions in question
and their precise language is unclear from the record.  A November 25,
1985 letter from Richard J. Gold, Divisional Deputy City Solicitor,
submitted by the State states that the motion asked "for a judicial
determination that it was in the best interests of the child not to be
returned home."  State ex. 6, 4th page.    However, in response to an
inquiry at the hearing as to "what the specific motion was," Judge
Nicholas A. Cipriani, who signed many of the placement agreements in
question, stated only that "[t]he motion was to approve the agreement of
voluntary placement."  TR 29.

15.  In further support of its position, the State also submitted nunc
pro tunc (NPT) orders in 22 cases, each of which stated in pertinent
part that, at the time the court signed the placement agreement, "it was
the practice of this Court to sign such Placement Agreements following a
court hearing if a judicial determination was made that it was in the
best interests of the child to be placed in foster care."  The NPT
orders further stated that "this Court can determine from its signature
on the Placement Agreement that such a determination was, in fact, made.
. ." and therefore ordered that effective on the date of the court's
signature on the placement agreement, "placement of the above named
child with DPW is in the best interest of the child."  Letter from Manne
to Horvath dated 10/23/90, enclosures.  Since we find that the court's
signature on the placement agreement necessarily represented a
determination that the child's removal from home was in the child's best
interest, however, the NPT orders are unnecessary.

16.  Nebraska is also distinguishable from the instant appeals on the
ground that the original court orders there were issued after ACF
clarified that a court order "must contain a statement to the effect
that continuation of residence at home is contrary to the welfare of the
child. . . ."  ACYF IM-85-25, dated August 14, 1985.  Moreover, prior to
the issuance of the original court orders, the State had suggested to
its judges that they address the required information in their orders.
Here, however, the original court orders were issued prior to October 1,
1978, before ACF interpreted the Act as requiring an explicit statement
in the court order regarding the contrary to the welfare determination.

17.  ACF also suggested prior to taking the disallowance in this case
that a written opinion was required in order to satisfy the requirement
for a judicial determination.  See letter from Manne to Garrett dated
1/30/91, p. 2.  Judge Cipriani testified, however, that no written
opinion was required in custody proceedings in which there was consent.
TR 20.  Although he noted that these cases were not strictly custody
proceedings in that they did not involve litigants arguing over a
child's custody, we see no reason why a different rule would have
applied here.

18.  ACF acknowledged, however, that it had found eligible payments in
some cases which fell within section 102(d)(1) and that the voluntary
placement agreements in those cases did not meet the criteria for such
agreements specified in section 472(f)(2).  TR 131-132.

19.  This documentation was provided for the following cases:  Sample
Case #190 (FY 1983), Sample Case #236 (FY 1983), Sample Case #5016 (FY
1985), Sample Case #5064 (FY 1985), Sample Case #5080 (FY 1985), Sample
Case #5022 (FY 1985), Sample Case #5111 (1985), Sample Case #5116 (FY
1985), Sample Case #5129 (FY 1985), Sample Case #5132 (FY 1985), Sample
Case #5142 (FY 1985), Sample Case #5192 (FY 1985), Sample Case #5206 (FY
1985), Sample Case #5269 (FY 1985), Sample Case #5322 (FY 1985), and
Sample Case #63 (FY 1986).  Letter from Manne to Garrett dated 2/1/91,
attachments.  The State later conceded that ACF properly found the
payments in Sample Case #5022 and #5142 erroneous on other grounds.
Letter from Manne to Garrett dated 9/13/91, p. 2.

20.  Additional documentation was submitted for Sample Case #190 and
Sample Case #5269.  As stated in our later discussion of these
individual cases, we conclude that this documentation was inadequate as
well and that the payments were erroneous.

21.  This figure does not include three cases which the State originally
contested but later conceded involved erroneous payments: Sample Case
#5022 (FY 1985), Sample Case #5142 (FY 1985) and Sample Case #5171 (FY
1985).  Letter from Manne to Garrett dated 9/13/91, pp. 2-3).  ACF
withdrew its finding of an erroneous payment in a fourth additional case
originally contested by the State, Sample Case #150 (FY 1986).  ACF
brief dated 9/14/90, p. 27.

22.  The Board asked whether the State had "any basis for disputing the
reviewer's findings for case #240 (FY 83) (referred to on p. 9 of the
Agency's brief dated 11/8/90)."  Letter to parties dated 8/16/91,
enclosure.

23.  Following the submission of post-hearing briefs, counsel for the
State forwarded to the Board an excerpt from the child's case record
which he alleged established the child's eligibility.  Letter from Manne
to Garrett dated 10/8/91.  The Board did not provide for any submissions
following the post-hearing briefs, and counsel did not advise the Board
that he intended to submit additional documentation in support of the
State's appeal.  While we would ordinarily give ACF an opportunity to
object to the inclusion of this document in the record, that is
unnecessary here since the document does not in fact aid the State's
case.  The excerpt consists of caseworker's notes which indicate that
the "District Office" verified that the case was an active AFDC case in
April 1971.  However, the notes further indicate that the District
Office later verified that assistance for the child's family was
discontinued on 7/1/71.  Id., attachment, p. 2.  Thus, the notes only
serve to confirm that the child was not deprived when she was removed
from home on 4/26/72.

24.  It appears, however, that even under the provision now in effect
(section 475(4)(B) of the Act), infants are not directly eligible for
foster care maintenance payments, although their needs are to be
considered in determining the amount of the payment made on behalf of
the mother.

25.  The State's letter refers to Sample Case #5269; however, it is
clear in context that the State intended to refer to Sample Case #5249.

26.  In view of this conclusion, we need not consider the other ground
for ACF's finding of an erroneous