California Department of Social Services, DAB No. 1003 (1988)

DEPARTMENTAL APPEALS BOARD

Department of Health and Human Services

SUBJECT: California Department DATE: December 12, 1988 of Social
Services Docket No. 88-61 Decision No. 1003

DECISION

The California Department of Social Services (State) appealed a
determination by the Acting Commissioner of the Social Security
Administration (SSA, Agency) setting the amount of federal fiscal
liability (FFL) to be credited to the State under the Supplemental
Security Income (SSI) program during the period October 1981 to March
1982. This case arises under the Federal/State supplementation
agreement (Agreement) whereby SSA administers supplementary payments to
SSI beneficiaries from state funds. Statistical sampling techniques are
used to determine FFL for the State supplementary payments SSA
erroneously made.

The Acting Commissioner determined that the Agreement had been modified
to permit SSA to change its original finding of "no error" in one
particular case to a finding to exclude that case from the statistical
sample. Because of the regression formula used to calculate the FFL,
the effect of changing the status of this one case would be to
significantly reduce the amount of FFL due the State. There is no
dispute that the original terms of the Agreement would not permit SSA to
change its finding.

In contrast to most of the appeals before this Board, this case does not
involve the calculation of amounts erroneously claimed for federal
participation. Rather, this appeal concerns the amount of State funds
erroneously paid by the Agency in the administration of a State program.
The central issue in this appeal is whether the Federal/State
supplementation agreement was modified, as SSA claimed, to allow the
change in the status of the case in question. For the reasons discussed
below, we find that the Agreement had not been modified to permit such
action by SSA and that, moreover, the circumstances here are
distinguishable from a previous situation where the Agency changed an
original finding. Accordingly, we reverse the Acting Commissioner's
determination.

Our decision is restricted to the narrow issues presented here
concerning whether the Agreement had been modified as suggested by the
Agency. Our decision does not preclude SSA from examining other factors
affecting the FFL credit, including the validity of any of the sampling
procedures, and the proper statistical treatment of the one particular
case given the large dollar amount associated with this case.

General Background

Pursuant to the 1973 Amendments to the Social Security Act, Pub. L. No.
93-66, section 212, which required states to supplement federal
Supplemental Security Income benefits, California makes State
supplementary payments (SSPs) to its needy, aged, blind and disabled
residents. Section 1616 of the Social Security Act (Act) provided that
states could avoid the burden of administering the program by entering
into an agreement with the Secretary of the Department of Health and
Human Services for federal administration of the supplementary payments
on behalf of the states. The State executed such a Federal/State
supplementation agreement for fiscal year 1982 (FY 82), wherein SSA, on
behalf of the Secretary, assumed responsibility for the administration
of California's SSPs. State Appeal File, Exhibit (Ex.) D. Under that
Agreement, SSA among other duties determined eligibility for and the
amount of SSPs and made the actual cash payments to recipients.

During FY 82, an essential element in determining the parties'
respective fiscal obligations arising from the Agency's administration
of the State's SSPs was FFL, a voluntary assumption by the Agency of
fiscal responsibility for a percentage of the State dollars it paid in
error as the result of incorrect SSP eligibility and payment amount
determinations. Pursuant to Article X of the Agreement, the Agency is
liable to the State for the dollar amount of erroneous payments that
exceed the national standard (4% for FY 82).

The procedures for determining the amount of erroneous payments were set
forth in Article V of the Agreement. SSA selected and reviewed a sample
of cases receiving SSPs in order to determine how accurately it was
paying SSPs. The rules and procedures for sample selection and review
were set forth in the SSA's Quality Assurance Manual (QAM), which was
incorporated by reference into the Agreement at Article V, Section F.
Agency personnel examined these selected cases in accordance with QAM
procedures to determine if eligibility factors had been properly
considered and if payment amounts were correct. Once such a
determination was made, a subject case would be moved from the "on-line
data base" (case pending review) to the "off-line regression data base"
(case review complete). When the case moved off-line, an "Original
Federal Finding" (OFF) would be generated indicating either a payment
error or no payment error (or as the Agency asserted it should have done
here--a finding that a case should be excluded from the sample). The
OFFs for the cases in SSA's sample review are then turned over to the
State. The percentage dollar amount of found payment errors is then
projected to the entire universe for the period.

The State then has the option to perform a review of a portion of the
federal sample. Article V, Section C of the Agreement. If the State
agrees with the SSA findings in all the cases in the subsample, SSA's
percentage overpayment computation will stand. If the State finds an
error in SSA's determinations in any case in the subsample, the
resulting difference is projected to the entire sample of cases by an
agreed regression formula.

In this appeal a State finding of "error" in a case after SSA had
reported "no error" resulted in a higher regressed error rate than would
have resulted if SSA had originally excluded the case from the sample
and then reinstated it in the sample as a State discovered payment
error.

Case Background

In its review of SSA's sample of cases, the State selected one case (the
Mueco case) where the State claimed SSA had mistakenly paid benefits and
the SSA review had not picked up this error.

The Mueco case involved an alien who received a combined SSI/SSP payment
during the sample period. According to the Agency's explanation of
events, SSA personnel erroneously moved the Mueco case from the on-line
data base to the off-line regression data base and an OFF of "no error"
for the case was transmitted to the State. The State had selected the
Mueco case for review and subsequently determined that the recipient
Mueco was ineligible for assistance and had therefore mistakenly
received an erroneous SSP of $1,808.50. As a result of the regression
formula the State had previously elected to use, the State's discovery
of the error in the Mueco case resulted in a significant increase in FFL
due the State.

In a proposed fiscal settlement agreement establishing net FFL for the
period at issue, however, SSA's Regional Commissioner stated that the
Mueco case should have been excluded from the federal sample by SSA case
reviewers and that the failure to do so was an error. State Appeal
File, Ex. A. The Regional Commissioner declared that, while the federal
data base is not permitted to be changed under the regression rules,
inclusion of the Mueco case would significantly distort the regressed
projection of federal error and result in extreme fiscal consequences to
SSA and a windfall to the State. Id., p. 2. Citing the parties'
treatment of a prior case as precedent (the Antolin case), the Regional
Commissioner proposed a FFL amount that reflected the treatment of the
Mueco case as a case that should have been excluded from the sample by
the federal reviewers and then reinstated as a State discovered error.

The State appealed the Regional Commissioner's decision to the Acting
Commissioner. State Appeal File, Ex. B. In confirming the Regional
Commissioner's decision, the Acting Commissioner declared that what had
transpired in the Antolin case had, in effect, modified the Agreement
between the State and SSA so as to permit dropping the Mueco case from
the sample.

In the Antolin case, SSA personnel made what was clearly a clerical
error reflecting an OFF of "payment error," which the State rereview
detected should have been "no error." The State failed, however, to
timely notify SSA of the mistake, thereby resulting in a combined
Federal/State finding of "error," causing FFL to be set initially at
$10.3 million for the relevant period. SSA later discovered its mistake
and requested the State, despite the expiration of the time period for
corrections, to submit a "difference letter" requesting a change in the
status of the Antolin case. The State did so, but later realized that,
because of the regression formula, the changing of the Antolin case to
"no error" resulted in reducing the FFL to $7.8 million. The State
thereupon protested the loss of $2.5 million in FFL, arguing that the
time limit for changing the Antolin case result had expired and thus the
correction of that case's status to "no error" should not have been used
to adjust the FFL.

In response to the State's objections, the SSA Regional Commissioner, in
a December 12, 1984 letter (Agency Appeal File, Ex. E), declared that
"the most equitable action to take is to change the Federal findings on
the FFL data base to a 'no error' entry . . . ." While acknowledging
that this action deviated from QAM guidelines, the Regional Commissioner
stated it was reasonable in light of the fact that the Antolin case was
part of the first regression sample for the SSI program and the
mechanics of inputting data were new to SSA and State personnel. It
should be noted that errors made by both parties contributed to the
problem in the Antolin case, though SSA's error initiated the problem.
(We note, moreover, that the State asserted that it was not in fact
required to submit a "difference letter" reflecting the State's finding
in the Antolin case.)

The effect of changing the Antolin OFF to "no error" raised FFL for the
period to $9.9 million. The State subsequently accepted the additional
FFL and signed the closing settlement agreement for that period.

It is undisputed that the way SSA handled the Antolin case contravened
section 9222(A)(1) of the QAM:

Any adjustment to original Federal findings impacts on the
statistical results of the sample when regression is applied.
Therefore, individual Federal findings may not be changed after
they are reported to the State even for transcription or other
clerical errors.

The Acting Commissioner found that SSA erred in not excluding the Mueco
case from the sample, and ruled that "SSA and California, by their
performance in the Antolin case, modified the terms of the Federal/State
Agreement to permit changing the Federal sample data base when it is the
product of an error and if not changed would create an inequitable
result to one party." State Appeal File, Ex. C, p. 7. The Acting
Commissioner stated that the December 14, 1984 letter from the Regional
Commissioner advising California of SSA's decision to change the data
base in the Antolin case and California's conduct subsequently accepting
the terms of that letter constituted "a mutually agreed on written
modification of the Agreement relieving the parties from strict
adherence to QAM section 9222(A)(1) when certain circumstances are met."
Id. The Acting Commissioner found that the facts of the Mueco case met
the conditions set forth in connection with the Antolin case and
confirmed the Regional Commissioner's decision that SSA's changing the
federal sample data base to exclude the Mueco case was justified.

Discussion

The Federal/State supplementation agreement contained the following
provision:

This Agreement may be modified at any time by a written
modification mutually agreed upon by the parties hereto.

Article XII, Section B.

The Agency's position is that the State's conduct in accepting SSA's
proposal to settle the dispute arising from the Antolin case modified
the Agreement so as to allow, under certain circumstances, the Agreement
to be freed from the provisions of the QAM, particularly section
9222(A)(1). This modification then permitted SSA to treat the Mueco
case as if the OFF had been to exclude it from the federal data base.
These circumstances are when: (1) an OFF is the product of an error;
and (2) if the OFF is not changed, an inequitable result will occur.

The State responded that the Agreement specifically states that it may
be altered only by "a written modification," a provision that both
parties understood to require a formal written instrument specifically
designed to modify the Agreement, as demonstrated by the fact that the
State and SSA have executed 13 formal modification agreements since the
signing of the Agreement in 1982. These 13 modification agreements
(State Appeal File, Ex. E) all bear the signature of a representative of
each party, along with a certification by the California Attorney
General that the signee for the State had the legal authority to make a
modification to the Agreement. This, according to the State,
demonstrates that "conduct" is not the recognized accepted means to
effectuate a prospective amendment to the Agreement. Furthermore, even
assuming that "conduct" could modify the Agreement, the State added that
it was never its intention in the Antolin case that the Agreement be so
modified; rather, the Antolin case should be viewed in its unique
circumstances and not as any type of precedent.

We find no merit whatsoever in the Agency's position concerning the
Agreement's modification. We find no support in the record for the
Agency's attempt to bootstrap the December 12, 1984 letter into a formal
modification of the Agreement. That letter makes no mention of any
proposed prospective amendment to the Agreement; its sole focus is a
solution to the dispute arising from the Antolin case. Contrary to what
the Acting Commissioner found, the letter did not formally set forth any
standards for rectifying an error in the sampling process. That letter
even acknowledged that SSA's proposal to change an OFF was a deviation
from QAM procedures. The only reasonable way to read that letter is as
a unique response to a particular set of circumstances.

Nor can anything reasonably be inferred from the State's "conduct" in
its acceptance of the resolution of the Antolin case dispute. There is
nothing whatsoever in the record of this appeal of any type of State
acknowledgment or response to the December 12, 1984 letter. All we have
been told is that the State accepted the increase of FFL to $9.9
million. The Agency would read this silence as consent to modification
of the Agreement. The State explained that it accepted the Regional
Commissioner's offer of increased FFL because it believed it was due
that FFL anyway. There is no indication anywhere that the State thought
it was agreeing to a modification of the Agreement in the Antolin case
by accepting the increased FFL. Clearly, there was no meeting of the
minds on this question.

The plain meaning of Article XII, Section B, of the Agreement is that
any modification of the Agreement must be in writing. As the State
pointed out, the parties adhered to that requirement on at least 13
occasions by means of signed "Modification to the Agreement" documents.
There clearly was a course of dealing between the parties on how the
Agreement could be modified. If SSA had meant to alter the Agreement,
as it now attempts to interpret it, SSA could have put forward such a
proposal in writing. But SSA did not do that in the December 12, 1984
letter. There is nothing in that letter to indicate that any permanent
modification of the Agreement was contemplated; there is no reason to
construe this letter as other than strictly an ad hoc solution to that
particular problem. In this regard, we find it significant that the
Regional Commissioner's letter proposing to change the OFF in the Mueco
case (State Appeal File, Ex. A) made no mention of the Agreement having
been modified. It was only later in the Acting Commissioner's
determination that SSA suggested that the Agreement had been modified in
any form by the Antolin case.

Moreover, the Agency's position conflicts with its own QAM and the
theory behind regression sampling. Under the Agency's interpretation,
any error in an OFF would be potential grounds for changing the federal
finding. Section 9222(A)(1) of the QAM anticipated that errors will
occur in sampling, but, because of the impact when regression is
applied, the OFFs should not be changed "even for transcription or other
clerical errors." If even clerical errors in OFFs were not to be
corrected, then the type of error that occurred in the Mueco case, as
discussed below, should also not be corrected. To do so would render
invalid the whole theory of regression sampling. Changing an OFF that
has already been reported to the State would affect the results of the
sample, calling into question the validity of the whole sample. The
State argued that, if the Mueco case should be treated as the Agency
suggested, the corrective procedure would be to take an entirely new
sample because the old sample had been irreparably tainted by the
treatment of the Mueco case.

Even if we were to accept the Agency's position that what transpired in
the Antolin case effectively modified the terms of the Agreement,
thereby relieving the parties from strict adherence to QAM section
9222(A)(1) under certain circumstances, we find that those circumstances
were not present in the Mueco case. The circumstances in the Antolin
case which, according to SSA, permitted an OFF to be changed were when:
(1) the OFF is the product of an error; and, (2) if not changed, an
inequitable result will occur.

Here the Acting Commissioner declared that the issuance of an OFF of "no
error" for the Mueco case was a mistake and the Mueco case should have
been excluded from the sample due to the presence of an unverifiable
eligibility factor (the sponsor's income). According to the Acting
Commissioner, this failure to exclude constituted an error that
satisfied the first Antolin criterion. As to "an inequitable result,"
the Acting Commissioner found that treatment of the Mueco case as a
State correction of a SSA finding of "no error" would result, because of
the regression methodology, in a net FFL for the six-month period
several times the average net FFL for all prior six-month periods.
Thus, SSA found that the Mueco case met the Antolin criteria and the
erroneous federal data base should be changed to prevent an inequitable
result to SSA.

We find that the errors that occurred in the Antolin and Mueco cases can
be distinguished. In its brief the Agency further defined an
Antolin-type error as a mistake of a "procedural or technical nature."
p. 31. In the Antolin case, the error was a simple clerical error; a
SSA employee failed to make a complete entry of the federal finding so
that the OFF reflected a finding different from that actually made.
Furthermore, the State contributed to the problem by not timely
notifying SSA of its finding. Moreover, the underlying SSA decision in
the Antolin case was substantively correct.

The error by the SSA reviewers in the Mueco case, however, was clearly
different; the record does not show that the OFF of "no error" was the
result of a clerical error. Rather SSA personnel, upon becoming aware
of problems concerning Mueco's sponsor, recalled their original OFF of
"no error" and conducted further investigations. When the further
investigations did not produce contradictory information, SSA personnel
made the apparently deliberate choice of closing the case and
rereleasing it to the State with a reissued finding of "no error."
State employees then checking other records (which, according to the
State, were also available to SSA--an assertion the Agency did not
refute) found that the reissued OFF of "no error" was incorrect. Thus,
the underlying SSA decision of "no error" in the Mueco case was
substantively wrong. Moreover, the change to a finding to exclude this
case, which SSA seeks, was also substantively wrong.

The only explanation given by the Agency for the rereleased finding of
"no error" was that this had occurred "through either an apparent
misunderstanding of [the relevant QAM section] or through some other
lapse." Agency Brief, p. 12. This explanation is speculative and does
not constitute evidence that the OFF was the result of a procedural
error.

Thus, we find that the error in the Mueco case was not a mistake of a
"procedural or technical nature," as the Agency characterized it, but
rather was the result of a deliberate decision-making process. Errors
of this type are, according to section 9222(A)(1), not to be changed if
the validity of the regression methodology is to be maintained.

As to an "inequitable result" if the Mueco case is not excluded from the
sample, it is undisputed that the retention of the Mueco case would mean
a much higher FFL credit for the State. That does not, however,
necessarily lead to an "inequitable result." There has been no
statistical testimony or evidence that it is inequitable, only that the
result is significantly higher. The Agency's assertion of an
inequitable result is just an assertion; it has not been proven.
Although the size of the FFL for this period could certainly be a basis
for further examination by the Agency of its correctness, there is no
tangible basis in this record for concluding that the result is in fact
inequitable.

Conclusion

For the reasons stated above, we find that SSA erred in its treatment of
the Mueco case. The Mueco case should be treated as a State finding of
"error" correcting a SSA finding of "no error." As stated earlier,
however, our decision does not preclude the Agency from examining other
factors affecting the FFL credit due the State.


________________________________ Judith A. Ballard

________________________________ Norval D. (John) Settle

________________________________ Cecilia Sparks Ford Presiding
Board