Department of Health and Human Services
Departmental Appeals Board
QUALITY CONTROL REVIEW PANEL
SUBJECT: Texas Department of
Human Services
Docket No. A-95-128
Decision No. QC85
DATE: July 31, 1995
DECISION
The Texas Department of Human Services appealed the May 3, 1995
quality control (QC) determination by the Acting Assistant
Regional Administrator (RA) of the Administration of Children and
Families (ACF) in state QC number CP0992 (federal QC number 306).
The RA determined that the assistance unit (AU) in this case was
ineligible for the $69.00 in Aid to Families with Dependent
Children (AFDC) assistance it received for the July 1994 review
month.
The AFDC payment at issue was prospectively budgeted pursuant to
a best
estimate that the local agency made in March 1994. The
dispute in this
case involves whether that best estimate was made
pursuant to permissible
state practice (PSP) and whether the
estimate was accurate.
We conclude that the March estimate was inaccurate because it did
not
reflect all facts that occurred during the time frame which
the local agency
used to determine the estimate. If the best
estimate is recalculated
pursuant to Texas PSP using such facts,
the client is prospectively
ineligible. Therefore, we uphold
ACF's finding that there was a $69
overpayment in this case.
APPLICABLE AUTHORITY
QC reviews are conducted against PSP. 45 C.F.R. � 205.42(b).
PSP is defined as a state's "written rules and policies that are
in
accordance with existing, approved State plan provisions." QC
Manual
(QCM), section 3130.
Eligibility for AFDC must be determined prospectively, i.e., a
local
agency estimates the AU's circumstances and income for
future months to
determine if the AU is eligible. Some states,
including Texas, also
determine the amount of assistance
prospectively. Prospective budgeting is based on a "best
estimate." The QCM defines a best estimate as one which:
(1) is based on a specific time frame which the State
uses to
determine the estimate; (2) accurately reflects
all facts that occurred
(whether known or unknown to the
State) during the time frame(s) which the
State uses to
determine the estimate; (3) is calculated correctly; and
(4) remains an accurate reflection of the likely
situation in the RM
because no change in circumstances
(as defined by PSP) has occurred since
the time frame
used for the estimate.
QCM, section 3420, at IV-6.
In a prospectively budgeted case, if there has been no change in
circumstances, the QC reviewer must determine the accuracy of the
best
estimate which was used as the basis for the review month's
payment.
QCM, section 3420 C, at IV-16. In order to determine
the accuracy of a
best estimate, the QC reviewer must:
(1) verify actual income and income-related
circumstances for the
time period which the local agency
should have used to establish the
estimate (i.e., use
all information that could have been known); (2)
recalculate the State's most recent best estimate,
following the State's
methodology (PSP) and (3) use all
appropriate disregards.
QCM, section 3420 C, at IV-17.
Therefore, if the reviewer determines that the local agency's
best
estimate was incorrectly calculated pursuant to PSP but no
change of
circumstances has occurred, the reviewer must
recalculate the payment based
on a best estimate calculated
pursuant to PSP. QCM, section 3420 C, at
IV-18.
In its Income Assistance Handbook, Texas set forth the following
PSP for
projecting income.
Anticipate income using the best available information.
If
income is ongoing, but the amounts fluctuate, it is
best to anticipate by
averaging income from past pay
periods. When using this method,
* verify at least four consecutive, recent pay amounts,
* evaluate other factors with the client to determine
if the
amounts verified are representative of future
amounts (don't include
unrepresentative amounts), and
* average the representative amounts to determine the
amount of income to project.
Exception: Use a different method to anticipate income
if
someone has a new job, or seasonal fluctuations or
expected changes (such as
changes in work hours or rate
of pay) cause too many past amounts to be
unrepresentative.
Different methods of anticipating future income are
* asking the employer for an estimate,
* using less than four pay periods,
* anticipating hours time rate of pay, or
* other methods.
Always document the reason for the method used.
Texas Income Assistance Handbook, A-784.
FACTS
The AU is this case is comprised of the client and her three
dependent
children. The AU received an AFDC payment of $69.00
for the review
month of July 1994. This payment was based on the
local agency's
estimate, made the preceding March, that the AU
would have monthly earned
income of $277.
The client began working on January 14, 1994, for C & D
Janitorial
and continued to work there through the review month.
On January 26,
1994, the client reported this employment to the
local agency and submitted
her first paystub. This was the only
paystub contained in the local
agency or Texas QC files at the
time of the federal re-review. The
client informed the local
agency that she would be employed 15 hours per
week and would be
paid at the rate of $4.25 per hour. The case file
reflected that
the employment change was processed on February 1, 1994 based
on
bi-weekly earnings of $127.50 (15 hours x $4.25 per hour x 2
weeks).
On March 15, 1994, the local agency made a full redetermination
of the
client's eligibility. For that redetermination, the case
record
indicates that the local agency spoke to the client's
supervisor concerning
the client's earnings. According to the
documentation in the case
file, the supervisor "verified the
employment and wages as std above."
The "above" list of wages
indicates the client had received $127.50 for the
pay periods of
February 2, 1994, February 16, 1994, and March 2, 1994 and
would
receive $127.50 for the pay periods March 16, March 30, and April
13, 1994. Generic Worksheet, at 7. Based on this information,
the local agency estimated future monthly earnings at $277
($127.50 x
2.17) per month.
On March 23, 1994, the local agency took action to reduce the
earned
income deduction from $30.00 plus one-third to $30.00
effective with the May
1994 payment. (Although the local agency
removed the one-third
disregard as of the May payment, both ACF
and Texas agree that the one-third
disregard actually expired as
of the June 1994 payment. This is
because the January earnings
would have been consumed by the $90.00 work
expense disregard so
that the $30.00 plus one-third disregard should have
started in
February rather than January and extended through June rather
than May.)
In the subsequent QC reviews, it was determined that the client's
actual
past employment circumstances differed from the
information the local agency
obtained and used in making the
estimate. When the federal QC reviewer
contacted the employer,
the employer said that the client had been hired at
$4.25 for 20
(not 15) hours a week, though hours could fluctuate. In
fact,
the client had the following actual income for the preceding pay
periods: $163.63 for February 2, 1994, $208.25 for February 16,
1994, $187.00 for March 2, 1994. Had the local agency used the
client's employment agreement of 20 hours per week or the
client's
actual amount of earnings as the basis of the estimate,
the client would
have been found, under Texas PSP, to be
ineligible for the July
payment.
From the record, it is not clear why the local agency failed to
obtain
accurate information about the client's income from the
supervisor.
One factor which could have contributed to the
incorrect information was
that, during the pay period the local
agency spoke to the supervisor, the
client was actually working
only 15 hours a week because of a disciplinary
action.
Therefore, the information that the client was working 15
hours
was correct, but only for that pay period. However, according to
the notes in the case record, the supervisor also said that 15
hours and
$127 per pay period were correct for all past pay
periods and for future pay
periods.
In its QC review, Texas found the case was correctly paid based
on its
determination that the local agency had estimated the AU's
earned income in
accordance with PSP at the March 15, 1994
redetermination and that there had
been no subsequent change of
circumstances which would require the estimate
to be
recalculated. Texas based its conclusion on a finding that,
under PSP, the local agency was authorized to base a best
estimate on
the employer's estimate of future earnings.
In its re-review, ACF determined that the case had been
incorrectly
paid. It based this determination on two grounds.
First,
"[b]ecause action was taken to alter the grant subsequent
to the March 15,
1994, best estimate, (i.e., elimination of the
1/3 disregard), an entire
best estimate . . . must be
recalculated at the time the earnings disregard
is altered." ACF
letter dated May 3, 1995. Second, ACF
maintained that the March
1994 estimate was incorrect because the
calculation was not done
pursuant to PSP and was based on incorrect or
insufficient
information supplied by the client's supervisor. Under
either
grounds, ACF maintained that the review month grant had to be
calculated pursuant to PSP on the basis of accurate February and
March
income information.
ANALYSIS
This case turns on the accuracy of the local agency's March
estimate. 1/ We conclude that the March estimate was
inaccurate because it was not based on correct information about
the
client's earnings. Under the QCM, when an estimate is
inaccurate and
there has been no change in circumstances, the
estimate must be recalculated
pursuant to PSP using the correct
information. Pursuant to this
recalculation, the AU was
ineligible for any assistance payment in July
1994. Below we
explain how we reached this conclusion.
For the following reasons, we conclude the March estimate was
inaccurate. The QCM defines a best estimate as one which
"accurately reflects all facts that occurred (whether known or
unknown
to the State) during the time frame(s) which the State
uses to determine the estimate." QCM, section 3420, at IV-6.
Further, the QCM provides:
In applying the "best estimate" concept, quality control
may have
to go beyond the information used by the local
agency. This is because
when a best estimate is
established, for purposes of calculating the AFDC
payment, it encompasses all information that could have
been known by
the agency regarding the assistance unit's
income and income-related
circumstances.
QCM, section 3420 A, at IV-7.
In order to determine the accuracy of a best estimate, a QC
reviewer is
directed to "verify actual income and income-related
circumstances for the
time period which the local agency should
have used to establish the
estimate (i.e., use all information
that could have been known)." QCM,
section 3420 C, at IV-17.
Evident in these definitions and instructions to the QC reviewer
is the
QC principle that QC, as it is presently structured,
"independently
establishes and verifies the facts about each
element of eligibility."
QCM, section 3020. That section also
provides:
. . . while QC is bound by the State's substantive
eligibility and
payment requirements, the QC review is
not limited to individual State
methods of verification.
In many instances, the QC review is a more
rigorous
examination of the circumstances relating to a case than
that
which was conducted by the eligibility worker.
Therefore, what is at issue in this case is whether "all facts
that
occurred (whether known or unknown to the State)" and which
"could have been
known" support a finding of eligibility under
Texas' PSP for a prospectively
budgeted case.
Applying these standards to the circumstances of this case, it is
clear
that the local agency's estimate was not based on "all
facts that
occurred." The local agency's information was that
the client was
employed for 15 hours a week when she was employed
20 hours and that, for
the three pay periods prior to the
estimate, the client had earned $127 a
pay period when the client
had actually earned $163.63, $208.25, and $187.00
for those pay
periods. Further, there is no indication in the record
that this
"actual income" was not information that "could have been known."
QCM, section 3420 C, at IV-17. The employer was a company which
issued paystubs which set out the pay period, the hours worked,
the rate
of pay, the gross pay, and the net pay. Had the local
agency obtained
such paystubs, it would have realized that the
client's and the supervisor's
representations that she was only
working 15 hours a week were not
accurate.
Texas argued that its best estimate is valid because its PSP
provides for
calculating an estimate on an employer's estimate of
future earnings and
that "the estimate obtained by the agency
reflected the client's employment
situation accurately at the
time of the last complete action." Appeal
letter, at 3.
Presumably, Texas means that during the week of March 15
when the
local agency contacted the supervisor, the client was in fact
working 15 hours. However, this coincidence of fact does not
negate the important ways in which the supervisor's apparent and
concurrent representations were incorrect or incomplete: (1) the
supervisor's representations about past earnings ($127 per pay
period)
were completely inaccurate -- the client's income had
fluctuated but it had
always been substantially over $127, and
(2) the reason the client was only
working 15 hours that week had
to do with a disciplinary action which was
not expected to
continue; she customarily worked 20 hours a week.
Texas' argument is not persuasive. While it is usually
appropriate
to defer to a state's construction of its PSP, Texas'
exclusive reliance on
the portion of its PSP authorizing best
estimates based on an employer's
prediction of the future is
unreasonable in this case.
2/ Although it is impossible to
know in advance whether an
employer's prediction will prove
accurate, a local agency's use of an
employer's prediction only
occurs after a local agency has information about
a client's
terms of employment and recent earnings in that particular
job.
In this case, Texas cannot justify its best estimate on the basis
of one aspect of the case record (the supervisor's prediction)
while
disregarding the fact that the local agency was materially
misinformed about
the client's earnings and that the local
agency's reliance on the
supervisor's prediction was predicated
on this misinformation.
Therefore, the cause of the error in
this case is that the local agency
based its estimate on
incorrect information. Texas' assertions that
the local agency
followed the procedures set out in its manual does
not change
the fact that the local agency's information did not "accurately
reflect all facts that occurred." QCM, section 3420, at IV-6.
Projecting the client's income on the basis of the average of her
three
representative paychecks or on the basis of a 20 hour work
week results in
countable earned income of $284.28 and $248.90,
respectively. Either
figure exceeds the payment standard of $226
for a family of four in
Texas. Therefore, this AU was not
eligible for AFDC assistance in July
1994.
CONCLUSION
For the foregoing reasons, we uphold ACF's determination that a
regular
ineligible discrepancy exists in this case in the amount
of $69 for the July
1994 review month.
Thomas D. Horvath
Leslie A. Sussan
Sara Anderson
* * * Footnotes * * *
1. We note that the
parties disagree as to whether the
elimination of the one-third disregard
subsequent to the March 15
estimate constituted a "change in circumstances"
which required a
new estimate. Section 3420 B. of the QCM makes it
very clear
that, for purposes of QC review, a change in a disregard is a
change of circumstances. ("Income and income-related
circumstances
involve . . . (4) all appropriate disregards."
QCM, section 3420 B.,
p. IV-8 - IV-9. Further, example 2.a. of
that section concerns the
expiration of the one-third disregard
as a change of circumstances.)
The local agency took the action
to eliminate the disregard on March 23,
1994 effective May 1,
1994. Both the initial federal review and the
RA's decision
based their decisions on the accuracy of the March estimate as
determined by information available as of March 23, 1994, thereby
treating the action of March 23 as the last best estimate.
Therefore, we will review this case on the basis of ACF's
conclusion
that the relevant facts concern those that could have
been known by the
local agency as of March 23, 1994.
2. ACF pointed out several ways in which it found that
the
local agency failed to follow PSP. These failures concerned
whether
the local agency should have used averaging the three
representative pay
periods rather than an employer's estimate and
the fact that the local
agency failed to document why it used the
estimate method. We need not
construe the Texas policy manual so
strictly here since use of a different
method to project income
would not have eliminated errors: i.e., if
the local agency had
applied the averaging method to the incorrect
information it had
gathered, there would still be a QC error. The
critical problem
here is that the local agency failed to obtain the correct
information rather than the particular method it applied to that
information.