DEPARTMENTAL GRANT APPEALS BOARD
Department of Health and Human Services
SUBJECT: Florida Department of Health and Rehabilitative Services
Docket No. 86-52
Audit Control No. 04-50511 Decision No. 821
DATE: January 9, 1987
DECISION
The Florida Department of Health and Rehabilitative Services
(State)
appealed a determination by the Office of Refugee Resettlement
(ORR)
disallowing $633,037 in stipend payments charged to project grant
funds
received by the State under the Cuban/Haitian Entrant Impact Aid
Program
(CHEIAP) for the period May 13, 1982 through June 30, 1983.
The
disallowance was based on an audit report which found that the State
had
improperly charged to the CHEIAP program costs in the
following
categories: $420,322 in stipends for ineligible participants;
$121,310
in stipends unsupported by daily classroom time and attendance
records;
$51,615 in stipends paid for holidays; and $39,790
in a "stipend
reimbursement overclaim." These amounts
include not only the direct
costs charged for the questioned stipends, but
also $64,270 disallowed
as related indirect costs.
The State disputed the findings regarding each
category of
the
disallowance, arguing that the stipend amounts
were in fact
allowable. In
addition, the State challenged the auditors' use of
projection from a
statistical sample as a basis for
determining
the amount of stipend payments unsupported by classroom time
and
attendance records. Finally, the State argued that there was no
basis
for disallowing related indirect costs.
For the reasons stated below, we uphold the disallowance in part
and
reverse in part, in an amount to be determined.
Background
Title V of the Refugee Education Assistance Act of 1980, Public
Law
96-422, provided at section 501(a)(1): "The President shall
exercise
authorities with respect to Cuban and Haitian entrants which
are
identical to the authorities which are exercised under chapter 2
of
title IV of the Immigration and Nationality Act [section 1521 et seq.
of
8 U.S.C.]." That Act included the authority under 8 U.S.C. 1522(c)
to
--
. . . make grants to, and enter into contracts with,
public or
private nonprofit agencies for projects
specifically designed --
(1) to assist refugees in obtaining the skills which
are
necessary for self-sufficiency . . .;
(2) to provide training in English where necessary . . .; and
(3) to provide, where specific needs have been
shown and
recognized by the Director, health (including mental
health)
services, social services, educational and other services.
On March 3, 1982, the ORR published a notice of availability of
funds
under this section. The notice announced "the availability of
funds and
award procedures for project grants for services to Cuban and
Haitian
entrants . . . in States and localities where specific needs exist
for
supplementation of currently available resources because of factors
such
as a high concentration of entrants." 47 Fed. Reg. 10908.
The notice
further explained:
The purpose of the grants is to provide additional
services to
entrants in areas where resources for
these purposes have been
unusually strained due to
factors such as especially large
concentrations of
entrants. Funding of these special projects is
intended to promote effective resettlement and to provide
needed
services to entrants while at the same time
helping to offset
extraordinary impacts or burdens
on State and local resources.
Id.
The program under which the funds were made available is called
the
Cuban/Haitian Entrant Impact Aid Program, or CHEIAP.
The Florida Department of Health and Rehabilitative Services, which
also
administers a State plan program providing basic cash assistance
to
Cuban/Haitian entrants, applied for CHEIAP funds and was
ultimately
awarded $31,025,000 for projects to provide services to
eligible
entrants during the period May 1982 through December 1983. One
of the
types of services to be provided under Florida's CHEIAP grant
was
employment services. The employment services project had
two
components: an intake/assessment component for evaluating an
entrant's
language ability and employment needs and an employment
training
component for actually providing training in English and
work-related
skills. Participants in the employment training component
could receive
stipends to meet their basic subsistence needs.
Florida needed to implement its CHEIAP program quickly.
Due to a
change in federal regulations, many
entrants who had been receiving cash
assistance under the basic Cuban/Haitian
Entrant Program would no longer
be eligible for that assistance if they had
been in the country for a
period of 18 months or longer. To implement
the employment services
project, the State entered into a contract with the
South Florida
Employment Training Consortium (SFETC), a local government body
which
had experience in running projects under the U.S. Department of
Labor
(DOL) program known as CETA. In turn, SFETC subcontracted with a
number
of other providers of employment services.
The HHS Office of Inspector General Office of Audit
performed an
audit of stipend costs claimed by SFETC
under the State's CHEIAP grant,
during the period May 13, 1982 through June
30, 1983. State's Ex. A.
Based on this audit, ORR disallowed direct
costs of stipends in the four
categories listed on page 1 above, as well as
the related indirect
costs. Below, we first discuss each category of
questioned direct costs
and then discuss the State's argument with respect to
indirect costs.
Discussion
I. Stipends for ineligible participants
Using a computer match of entrant Social Security
numbers with
the State's wage
and earning records, unemployment compensation benefit
records, and cash
assistance payment records, the auditors found that
the State made
stipend payments of
$377,648 to entrants who
were either
employed or simultaneously
receiving income from
other sources. ORR disallowed this amount on the
basis that the receipt
of income rendered the entrants ineligible to receive
the employment
services stipends. ORR based its decision on its view
that State
regulations and the terms of
the grant required the State to redetermine
eligibility on a continuing basis
in order for the stipends to be
allowable.
State regulations on eligibility for the employment services
program
stated that such services would be provided to individuals who
qualified
as "entrants" and who, in addition, met the
following criteria,
among others:
(b) The individual must have reached the time limitation
for
cash and medical assistance from CHEP [i.e., the
Cuban/Haitian
Entrant Program].
(c) The individual must not be employed nor receiving
any
income, as evidenced by a statement signed by the individual.
Florida Administrative
Code,
section 10C-30.08(1),
State's
Ex. E.
ORR did not deny that, for each of the participants
in
the
employment services program, SFETC obtained a signed statement from
the
individual that the individual was not employed and had no income,
nor
did the auditors find that any
of the individuals gave false
information at the
time that eligibility was
determined.
Rather, ORR relied on its view that
an individual became ineligible
for a stipend whenever that individual began
receiving any other income
and that the State's interpretation of its
regulation (as requiring only
signed statements at the start of the program)
was inconsistent with the
State's interpretation of regulations in the
medical services program
operated under CHEIAP.
The State explained that it did not require
redeterminations of
eligibility for the
employment services program every six months, as it
did for the medical
services program, because the
employment
services were offered for a period of only six months at
most. The
State further explained that the purpose of the stipend
payments (in
addition to providing basic subsistence) was to provide
incentive to the
participant to complete the employment training, so it made
sense not to
require termination of the stipend based on later receipt of
income.
The State also pointed out that it had to implement the program
very
quickly, using a number of service providers, and that, in that
context,
it was reasonable of the State not to adopt an overly
burdensome
eligibility process. When ORR alleged that the State could
have easily
performed a computer match such as the auditors had, the State
offered
evidence that the income would not have shown up on
the computer
program until several months after it had been received.
For the reasons stated below, we uphold the disallowance of stipends
paid
to individuals who received cash assistance under CHEP and reverse
the
disallowance of stipends to individuals with other forms of income.
ORR
should determine these amounts from the audit workpapers.
ORR alleged, and the State did not deny, that some of the income which
the
auditors found had been received by participants was cash assistance
under
the CHEP program, paid for periods during which the individuals
were
receiving stipends. It necessarily follows that these individuals
had
not exhausted their 18-month eligibility for CHEP assistance.
The
wording of section
10C-30.08(1)(b) is mandatory and
unqualified, i.e., the individual must have
reached the time limit
for
CHEP
assistance, and is based on the grant application,
which described the
employment services as being for "those entrants
losing cash assistance due
to the 18 month termination policy." ORR's
Ex. H.
The State apparently implemented this by providing to
SFETC a list
of individuals who had been
terminated from the CHEP program as a result
of the 18-month limit.
SFETC said that it nonetheless provided
employment services to some
individuals who were not on the State's list
because of the possibility that
some individuals had stopped receiving
CHEP assistance earlier. Even if
this might have been reasonable in
some instances, however,
we do not think it reasonable that neither
the State nor SFETC implemented
any procedure to verify that individuals
not on the list had met this
condition. Indeed, the intake documents
used by SFETC indicate that
even the minimal step of asking the
applicant about this requirement was not
taken. State's letter of
October 13, 1986, Exs. A-C.
Thus, we uphold the disallowance to the extent that it
relates to
individuals who had not met the condition of
having exhausted their
eligibility for CHEP assistance.
On the other hand, we conclude that, with respect to other kinds
of
income, the State's interpretation of its own regulation is a
reasonable
one, consistent with the terms of the grant. We base this
conclusion on
the following:
o Unlike the wording of the Florida Administrative Code
provision
related to the medical services program, which states all of
the
eligibility requirements as conditions the individual "must" or
"must
not" meet (see section 10C.30.03), the provision on employment
services
at section 10C.30.08(1)(c) specifies the particular "evidence"
which
will be deemed acceptable in determining whether an applicant
is
employed or receiving income, i.e., a signed statement from
the
applicant.
o The State's interpretation of 10C-30.08(1)(c) is consistent
with
Florida's grant application. The corresponding section of the
grant
application provides that services should be extended to
eligible
entrants who are "currently unemployed with no income." ORR's
Ex. H, p.
16 (emphasis added).
o ORR specifically asked the State, with respect to eligibility
for
health services, to justify the State's approach if it intended not
to
redetermine eligibility for the one-year duration of the project.
The
State's response was: "Redetermination of eligibility will
be
determined on a six months basis, pursuant to 42 CFR 435.831."
ORR's
Ex. I, question 4. No such question or response was made for
the
employment services part of the project.
o Another indication in the grant application of
the State's
different
approach to the two programs is that eligibility
determinations for the two
types of services were performed in a
different manner. For health
services, periodic redeterminations of
eligibility would appear
to be easier since they were made by the
local county
health departments. ORR's Ex. I, question 5. For
employment
services, eligibility was to be determined by the "provider,"
the actual
entity which provided
the
training. ORR's Ex. H,
p. 18.
o Even if a six-month redetermination rule at 42 CFR 435.831
should
have been applied to employment services, it would not have made
a
difference here since it is undisputed that the employment services
were
designed to be offered for a period of only six months. See ORR's
Ex.
H, p. 16. This also appears to be a logical explanation for
a
difference between the two programs and underscores to us
the
reasonableness of the State's interpretation.
o It appears that determining eligibility for employment
services
stipends in the manner accepted by the State is consistent with
the
goals of the program. The program was intended to provide a
quick
response to an immediate need. One purpose of the stipends was
to
provide an incentive to participants to remain in the program. The
need
for this incentive would remain even if the recipient was
receiving
funds from some other source, so long as the recipient had not met
the
ultimate goal of the services, which was to obtain skills necessary
for
self-sufficiency. Also, although the stipend was supposed to
provide
basic subsistence, it appears to have been only about half the
amount
paid by other on-the-job training programs. See ORR's Ex. I,
question
10. Thus, it appears reasonable that, rather than setting up
a
redetermination process, the State might take some risk that a
person
who was eligible initially might obtain some additional income
from
another source.
We also conclude that ORR's reliance on 45 CFR 74.170
and the
cost principles in
Office of Management and Budget (OMB) Circular A-87
is misplaced here.
These provisions provide generally that a state can
charge only allowable
costs to federal funds and must
document
the allowability of
costs which are charged to
federal funds,
but this begs the question of whether these
specific costs are
allowable. OMB Circular A-87 does not contain any
provision specifying
the type of documentation necessary for establishing
eligibility of
recipients in a services project. Here, the State did
document that the
individuals were currently receiving no other income at the
time of
applying for employment services. ORR has not shown that the
State was
required to redetermine eligibility on a continuing basis as
ORR
alleged.
Accordingly, we reverse the disallowance of stipends paid to
individuals
who had exhausted their 18-month eligibility
for CHEP and who had
no other income at the
start of the program, as evidenced by their
signed
statement. As stated above, we uphold the disallowance relating
to
those individuals who had not met the condition of having exhausted
their
eligibility for CHEP assistance.
II. Daily classroom time and attendance records
The auditors found, based on a review of a sample of stipend
payments,
that the State improperly claimed reimbursement
for at least
$108,994 in stipends which were not
supported by subcontractors' daily
classroom time and attendance
records. The State challenged the use
of a
sample to project a disallowed amount and also disputed some of the
findings
on individual sample items. In addition, the State pointed out
that
$59,999 in stipends had been paid from non-federal
cash
contributions, arguing that this amount should have been used to
offset
the disallowance for unsupported stipends.
We address each of these issues below. We uphold use of
a sample
projection under the
circumstances here, but reverse the ORR findings on
some (but not all) of the
disputed sample items. We further conclude
that the $59,999 cannot be
used to offset the disallowed amount, but
may reduce
that amount if it was improperly included in the universe
based on which that
disallowance was projected.
Use of projection from a sample
Each of SFETC's subcontractors was required to maintain a payroll
and
attendance record (PAR) for each program participant. The PAR
was
generally prepared by counselor supervisors and used as a basis
for
determining the amount of the stipend to be paid to the
participant. In
addition, classroom instructors recorded time and
attendance (although
the methods for doing this varied). The auditors
found that, in
general, the hours indicated on the PARs did support the
amount of the
stipend check, but that there were discrepancies between the
attendance
hours on the PARs and the instructors' daily classroom time
and
attendance records. To determine the amount of stipends unsupported
by
the instructors' records, the auditors used a sampling process,
which
they described as follows:
. . . we used a computer application to
obtain a
random statistical sample. We examined 600 transactions
from
a population of 39,753
checks with a total value
of
$3,281,981. The population of transactions that we
sampled
pertained to the period August 1, 1982 through July 3,
1983.
The sample disclosed that 155 of the 600 transactions were
not
appropriate charges. Projecting the results of
the
statistical sample over the population, we are 95
percent
confident that, for August 1, 1982 through July 3,
1983,
stipends of at least $108,994 and$12,316 in related
indirect
costs were inappropriately charged to the CHEIAP
program.
State's Ex. B (Audit report), p. 10.
The State did not contest the validity of the statistical sampling
method
used by the auditors but contested the
use of projection
from the sample
as a basis for determining a disallowance on the ground
that it was unfair to
the State. The State first seemed to be saying
that it could not
adequately contest the audit findings if the full
amount of the projected
disallowance was not related to individually
identified transactions.
In response, ORR cited to various court cases
upholding use of projection
from a sample as a basis for identifying an
amount of unallowable
costs. ORR pointed
out that the State
was required only to refute the auditors' findings on the
questioned
sample items; if the State
could show that a dollar was
improperly
disallowed in the sample, then
the projected disallowance would be
reduced accordingly. Subsequently,
the State clarified that the reason
that
it thought that use of a sample
projection was unfair here
was that SFETC cannot recover disallowed amounts
from its
subcontractors without being able to document what specific costs
are
unallowable. According to the State, if specific data is not
provided
to support the entire disallowance, the local political
jurisdictions
that make up the SFETC must unfairly use local taxpayers'
monies to
repay disallowed costs. The State also argued that DOL does
not project
disallowances from samples when auditing similar employment and
training
programs and, therefore, it was unfair of HHS to use different
audit
procedures.
The Board and the courts have upheld the use of statistical
sampling
evidence as a basis for a disallowance, where claim by claim review
is
not feasible. See, e.g., Georgia v. Califano, 404 F.Supp.
(N.D.Ga.
1977); Rosado v. Wyman, 322 F. Supp. 501 (D.N.J. 1977); Ohio
Department
of Public Welfare,
Decision No. 226, October 30,
1981. Here, the
auditors reasonably employed a statistical sampling
method to determine
the amount of unallowable costs in a
universe of 39,753 stipend
checks. Moreover, the statistical validity of the
method used is not
challenged by the State. It is well-established that
a valid
statistical sampling methodology produces an extrapolated result
which,
with a high degree of probability, is as accurate as if a
full
examination of sampled items was done; indeed, it may be even
more
accurate because of the possibility of errors which can occur
in
examining a large number of items rather than a sample.
The State's arguments concerning fairness do not
establish that
the State was
hampered in its ability to dispute the disallowance; as
ORR pointed out, by
disputing individual sample items found to be
unallowable the State
effectively disputes the projected amount as well.
Rather, what the State
perceives as unfair is the fact that it (or its
contractor SFETC) may have to
bear the whole burden of the disallowance
if it cannot document the precise
items of unallowable cost for each
subcontractor. This argument ignores
the fact, however, that the State
has assumed the responsibility for
accounting for federal funds. With
respect to the federal government,
the State is liable for
misspent funds,
regardless of whether it can recover them from a
contractor or whether that
contractor can recover from subcontractors.
Moreover, while the State (or
SFETC) may not be able to recover from
the
subcontractors based on the federal audit because of use of the
sampling
technique, the State (or SFETC) still has the option of
performing its own
audits of the subcontractors to identify and recover
specific unallowable
costs.
We also reject the notion that this Department should be precluded
from
using a valid audit technique simply because DOL does not use it
for
similar types of programs. The State implied that it was prejudiced
by
not knowing in advance that statistical sampling might be used by
HHS
auditors, but did not explain how. In our view, advance knowledge
of
what audit technique would be used was not necessary in order for
the
State to operate its program.
Thus, we uphold ORR's use of a projection from a sample to calculate
the
unallowable costs related to unsupported stipend payments.
Disputed sample items
The auditors originally questioned sample stipend payments
totaling
$4,277.90. The State provided additional documentation to
challenge
findings with respect to $1,548.73 of that amount. The
auditors
accepted the documentation with respect to $1,039.00 (and reduced
the
projected disallowance accordingly) but rejected the documentation
for
$510.12. (The State never provided documentation to dispute the
audit
findings on the remaining $2,729.17 of questioned stipend
payments.)
During Board proceedings, ORR clarified why the auditors had not
accepted
the State's documentation related to the $510.12 (involving
payments to
program participants). The State conceded that the auditors
had
correctly rejected the State's documentation for five participants
for
stipends totaling $61.50. Documentation for seven of the
participants
for a total of $70.00 in stipends showed that the
questioned hours of pay
were for holiday leave. We discuss this issue
in the next section of
this decision and conclude that, contrary to
ORR's position, stipends could
properly be paid while participants were
on holiday leave. With respect
to the remaining stipend payments in
dispute, we have set out individual
findings in an appendix to this
decision.
While we have evaluated the documentation for each
participant
individually, there are certain general considerations which have
guided
our review. In our opinion, ORR's review was flawed because
ORR
operated on the assumption that the classroom instructors'
records
provided a more reliable basis than the PARs for determining
the
appropriate amount of a stipend. The State asserted
without
contradiction, however, that the instructors' records did not
account
for other than regular classroom hours, such as counseling
and
employability skills training hours, or for excused absences.
See
State's Response to Order to Develop the Record, Ex. A, p. 2. For
the
most part, it appears that ORR did give the State an opportunity to
show
that an absence from the classroom was excused or was for
other
reimbursable training. But we found that ORR did not give the
PARs
sufficient weight in situations where there was no actual
discrepancy
between the PAR and the instructor's record (as the auditors
said), but
where the instructor's record was simply incomplete. In
those
situations, the PAR is some evidence that the counseling
supervisor,
responsible for determining the appropriate amount of the
stipend, made
a contemporaneous determination that the recipient had in fact
attended
class in spite of the incompleteness of the instructor's
record. Absent
affirmative evidence that the participant did not in
fact attend, the
unrefuted reliability of the PAR demonstrates to us that the
stipend
should be considered allowable.
On the other hand, for some of the sample items, it appears that
the
person who prepared the PAR made a mistake either in calculating
the
hours shown on the instructors' records or in transferring the
recorded
hours of classroom attendance to the PAR. In these cases, we
agree with
ORR that documentation not provided to the auditors, but which
was
produced after the fact and which conflicts with documentation
provided
to the auditors, is unreliable and, therefore, should be
rejected.
Based on our analysis of the record here and in the Appendix, in light
of
these considerations, we uphold unconditionally the findings for 9 of
the
contested participants (totaling $118.12) and reverse the findings
for 17 of
the participants (totaling $348.25). For one of the
participants (I.,
Celeste, for a stipend of $43.75), we uphold the
disallowance unless the
State submits to ORR certain information within
30 days of receipt of the
decision. (We note, of course, that the State
presented no
documentation with regard to an additional $2,729.17 in
stipends, so these
findings are upheld without discussion.)
The offset issue
The State contended that $59,999 in cash contributions made by
Catholic
Community Services (CCS) was used to pay stipends to recipients
of
employment services and therefore this amount should have been used
to
offset the disallowance for unsupported stipends. ORR took the
position
that the $59,999 should not simply be offset against the
disallowed
amount since these CCS funds were restricted to stipends paid by
one
subcontractor of SFETC.
In an Order to Develop the Record issued in this case,
the Board
noted that ORR appeared
to be correct that the proposed offset would not
be proper, but noted that,
if some of the audit exceptions in
the
sample were for stipends paid out of funds donated by CCS, then it
would
appear that those sample exceptions would not be instances of
misspent
CHEIAP funds and therefore those sample payments should not have
been
included in calculating the disallowance.
In commenting on this,
the State did not deny that the use of CCS funds was
restricted to
stipends paid by one
subcontractor of SFETC. Further, the parties
agreed that none of the
individual sample payments found to
be
unsupported related to that subcontractor. The State, however,
argued
that, without the donation from
CCS, fewer entrants would have
been
served or more costs would have
been charged to the CHEIAP
program. The State argued, essentially, that
it was unfair of ORR to
ignore this sizeable cash contribution, which was
used to enhance and
enlarge a program established because of the significant
problems caused
by the sudden infusion into South Florida of a high number
of
non-English speaking, non-acculturated individuals.
We sympathize with the State about the problems it faced
due to the
impact of the Cuban/Haitian
Entrants. But the State's difficulties do
not provide a basis for using
a cash contribution restricted by its own
terms to costs incurred by a
specific subcontractor to cover unallowable
costs incurred by other
subcontractors, which would be the effect of an
offset. The State's
argument does point out one possible flaw in the
disallowance, however.
If in projecting the disallowance to the
universe of stipend payments, the
auditors used all stipend payments,
including those paid for by the CCS
contribution, rather than using a
universe consisting solely of stipend
payments actually charged to
CHEIAP funds, the disallowance amount is
overstated. Thus, the State
should have an opportunity to show that the
universe of stipend payments
made with CHEIAP funds was less than the amount
used by the auditors in
calculating the disallowance. If so, the
disallowance should be
adjusted accordingly.
In summary, we uphold ORR's use of projection from a statistical sample
as
a basis for disallowance, but reverse ORR's findings on
17
individual sample items. ORR should recalculate the
disallowance
accordingly, also taking into account any adjustments to the
universe
necessary so that it includes only stipend payments charged to
CHEIAP
funds.
III. Stipends paid for holidays
ORR disallowed $46,375 for the payment of stipends to individuals
enrolled
in SFETC's employment training program for those days
designated as
"holidays" by SFETC. ORR argued generally that payments
for holiday
leave were unallowable under the applicable cost principles
and that ORR had
no knowledge of SFETC's policy to provide stipends for
holiday leave.
We conclude that the State's payment of stipends for holiday leave
was
fully reasonable under the applicable cost principles and in
the
circumstances of this case. As explained below, we find that
the
particular provision of the cost principles cited by ORR is not
directly
applicable to the circumstances here and the State has
adequately
demonstrated that this policy of compensating for holiday leave
fully
accorded with purposes of the CHEIAP program.
SFETC's written policy regarding holiday leave provided as follows:
Absence for good cause shall be defined as follows:
1) Holidays which are limited to Christmas, New
Year's
Day, Memorial Day, Washington's
Birthday, the Fourth
of July, Labor Day,
Veteran's Day, Columbus Day,
Thanksgiving, the
day after Thanksgiving and other
holidays
specifically approved by SFETC.
* * * *
ORR's Ex. AA (SFETC's Sub-Grantee Policies and Procedures Manual),
third
page. The State explained that the reason for the holiday leave
policy
was to encourage participation and interest in the program by Cuban
and
Haitian entrants. The policy helped to ease the cultural barriers
faced
by the entrants, by affording them many of the same holidays enjoyed
by
the entrants in their own culture.
ORR made several objections to the policy of holiday leave, which
we
consider in turn below.
ORR cited OMB Circular A-87, establishing cost principles for state
and
local governments. Under the standard for "employee fringe
benefits,"
the Circular provides that "holiday leave" may be allowable as
an
employee fringe benefit for individuals who receive compensation
for
services rendered. See OMB Cir. A-87, Att. B., section
B(13)(a). ORR
argued that, since participants in the CHEIAP program
were not
"rendering a service," but were rather receiving a service, they
should
not be eligible for holiday leave under this cost principle.
We do not think that the fact that OMB Circular A-87
specifically
addresses "holiday leave" only in the context of providing
compensation
for services rendered means that the cost of holiday leave
is
unallowable in the situation of participation in a training
program.
OMB Circular A-87 nowhere precludes payment of
holiday leave for
participation in a training program, and we do not infer an
intent to
preclude payment in this situation merely because holiday leave
was
defined only in a different context.
Further, the fact that holiday leave is typically
extended to
employees (as reflected
in OMB Circular A-87's provision) would actually
appear to support the
State's view that holiday leave is appropriate
here, in the absence of some
evidence to the contrary. The stipends
were
intended as subsistence payments and it appears reasonable that the
trainees,
like employees, would expect to be paid for holidays. This
would appear
to be especially so for a training program that prepared
participants for
employment.
ORR also argued that, while the State may have had a written
policy
authorizing payments for holiday leave, ORR was not aware of such
a
policy and did not approve it.
Whether ORR had actual knowledge of the State's policy to
provide holiday
leave also does not appear relevant, if the policy was
otherwise legal and
proper. ORR cited no authority for the proposition
that ORR must know
of or approve all policies implemented by a state
under the CHEIAP
program. Moreover, even if ORR had no actual knowledge
of the policy,
the State, by formally establishing the policy in written
form, allowed ORR
at any time to examine this and other specific
policies, if ORR had
wanted.
ORR also relied on a State administrative rule pertaining to
compensation
for training programs, which states:
Individuals enrolled in training and employment
services may
receive $1.75 per hour for each
hour of attendance in the
employment services
program.
Florida Administrative Rule 10C-30.08(3). ORR maintained that,
since
participants were not "attending" the training
program when they
received credit for holiday
leave, this rule precluded payments for this
purpose.
In our view, this rule was intended to establish the applicable
hourly
rate of reimbursement and was not designed to describe all
the
particular circumstances in which payment could be made. In any
event,
the rule does not even refer to actual attendance in an
employment
training class, but rather refers more generally to attendance in
the
"program." As explained above, we find it reasonable here for the
State
to have regarded the scope of the program as including consideration
for
holidays.
ORR also cited several general provisions of OMB Circular A-87
which
require that for a cost to be allowable, it must be: 1.
"necessary and
reasonable for proper and efficient administration of the
grant
programs," and 2. "be accorded consistent treatment through
application
of generally accepted accounting principles appropriate to
the
circumstances." OMB Cir. A-87, Att. A, sections C(1)(a) and
(d). ORR
argued that the policy of compensating
for holiday leave was
implemented "inconsistently" by the State, since not
all of SFETC's
subcontractors paid participants for holiday leave. ORR
further argued
that, since the program nonetheless worked smoothly for
those
subcontractors who did not provide holiday leave, the State had
not
demonstrated the costs associated with holiday leave to be
"reasonable
and necessary." Id., p. 13.
We do not find that the State treated costs here inconsistently,
as
prohibited under OMB Circular A-87. "Inconsistent treatment of
costs"
generally refers to a grantee's treating federal funds in a
manner
inconsistent with the grantee's treatment of its own or other
funding.
This is different from the instant situation, where ORR has
not
demonstrated that it was prejudiced by the fact that some
subcontractors
chose to pay for holiday leave and others did not. The
State, indeed,
was entirely consistent in its providing a policy authorizing
the
payment of holiday leave, which subcontractors in their discretion
could
choose to implement.
ORR also specifically objected to the granting of holiday leave
for
holidays other than the 10 holidays specifically mentioned in the
SFETC
Policies and Procedures Manual, such as Martin Luther King's
birthday.
The decision's appendix lists two individuals for whom ORR
argued
improperly received stipends on this basis. We conclude
that
subcontractors could extend holiday leave for these other
holidays. The
Manual defines an absence for good cause to include
"other holidays
specifically approved by SFETC." ORR did not argue that
these other
holidays were not approved by
SFETC. Thus, we have no basis for
concluding that payment for these
days was inconsistent with SFETC's
policy, which we have decided above was a
reasonable one.
Accordingly, we reverse the disallowance for stipends paid for
holiday
leave. Nothing here, of course, precludes ORR from establishing
a
different prospective policy on stipends for holiday leave.
IV. Stipend reimbursement overclaim
ORR disallowed $35,750 for employment training stipends
which ORR
alleged were not adequately supported by
payroll accounting records.
SFETC obtained its payroll services from the Dade
County Office of
Computer Services and Information Systems (OCSIS). The
auditors
compared the amount of checks
issued to the program participants
for
stipend payments with "SFETC's cost report for stipends paid" in
order to
calculate the $35,750 disallowance. State's Ex. B (Audit
Report), p.
14. As explained below, we reverse this part of the
disallowance.
The State explained during the appeal that the "cost report" examined
by
the auditors was actually one of several data bases used by OCSIS,
and
argued that, if the more appropriate data base were instead
considered
when comparing the amount of stipend checks issued, the
$35,750
discrepancy would be reconciled.
ORR did not deny the existence of
these different data
bases and, as explained below, we conclude
that
ORR disallowed the $35,750 because of confusion as to the
appropriate data
base to consider.
The State explained that the data base examined by the auditors was
the
"payroll data and earnings histories." State's Response to Order
to
Develop the Record, Ex. A, p. 2. This data base is maintained for
tax
purposes, according to the State. The State asserted that the data
base
maintained by OCSIS which would accurately reflect the amount
of
stipends paid was the system used for financial, rather than
tax,
accounting, called the Financial Accounting Management
Information
System (FAMIS). The primary
reason for discrepancies between the
two systems is "adjustments" made "after
the payroll has been issued."
State's Response to Order to Develop the
Record, Ex. A, p. 3. When a
participant in the program is paid under
"the wrong code," "adjustments
are made manually to FAMIS and are never
reflected in the payroll
system," i.e., the payroll data and earnings
histories data base,
maintained for tax purposes. Id. In other
words, any computer or
administrative errors reflected in the payroll and
earnings histories
data base would be corrected manually on FAMIS, to
accurately reflect
the amounts of stipends paid.
The Board inquired whether the State had evidence that use
of the
FAMIS system would reconcile the discrepancy
discovered by the auditors.
The State first noted that SFETC had presented to
the auditors "original
hard copy documentation" of the payroll transfers to
all participants in
the program for the period audited. ORR found no
discrepancy from this
evidence, other than to
slightly increase the amount for which the
State could claim because of a
technical error. State's Response to
Order to Develop
the Record, Ex. A, p. 3.
In order to further demonstrate the point that ORR relied on the
wrong
data base, the State analyzed the treatment by FAMIS of one of the
two
payroll divisions which together accounted for most of the
$35,750
discrepancy. For this payroll division, the James E. Scott
Community
Association, Inc. (JESCA), the State compared the total amount
recorded
under the earnings history data base with an amount which was
extracted
from the FAMIS system. On this basis, the State calculated
a
discrepancy of $11,006.65. See State's Post-Conference Brief, p.
13;
State's Exs. H-K.
ORR presented no evidence or argument to refute this demonstration of
the
difference between the systems, or why the data base which ORR
relied on
should be considered more accurate. ORR did submit that it
did not have
an opportunity to specifically review the data presented by
the State (the
calculation of the $11,006.65 discrepancy involved the
tabulation of earnings
statements for each of 105 participants), and
noted that, in any event, the
State did not reconcile the entire $35,750
discrepancy, but only some part of
it.
We do not agree with ORR's implication that the
State
must
reconcile the entire $35,750 discrepancy in order to demonstrate
its
general point that ORR examined the wrong data base. The State
appears
correct that use of the FAMIS system would corroborate the amount
of
stipends actually paid. If ORR wants to more thoroughly analyze
the
data presented by the State and use of the FAMIS system, this
decision
will not prevent it from doing so. We reverse the disallowance
based on
ORR's use of an earnings history data base, which
we have found on
the record of this case to be the wrong data base to
use. If ORR
examines the use of the FAMIS system and nonetheless
determines that the
State cannot account for the total amount of stipends
claimed under the
CHEIAP program, ORR may disallow such amounts on this
basis, and the
State may again appeal to the Board.
V. Related indirect costs
In addition to the $633,037 disallowance under the four
categories
discussed above, ORR disallowed $64,270 in related indirect
costs. The
State initially objected to the disallowance of indirect
costs as
"another arbitrary and capricious action" which was "without any
legal
authority." State's Opening Brief, p. 16. As briefly
explained below,
ORR may disallow indirect costs claimed based on direct
costs found to
be unallowable.
SFETC had negotiated an indirect cost rate agreement with the
U.S.
Department of Labor, which authorized SFETC to claim indirect
costs
associated with any direct costs at the rate of 11.3 percent.
Agency's
Ex. L. Under the agreement, the 11.3 percent rate was
applicable to all
grant programs with the federal government to which OMB
Circular A-87
applies, which includes the CHEIAP program. On the basis
of this
agreement, the State claimed indirect costs in operation of
the
employment services project at the rate of 11.3 percent of direct
costs.
A grantee's entitlement to reimbursement for indirect costs is premised
on
its entitlement to reimbursement for the direct costs with which
those
indirect costs are associated. In other words, only that portion
of
indirect costs allocable to allowable direct costs may be charged to
federal
funds. See, e.g., OMB Cir. A-87, Att. A, section D(1). Thus,
we
conclude that, to the extent the stipend payments which were claimed
as
direct costs of the CHEIAP grant were unallowable, the associated
indirect
costs are also unallowable.
Conclusion
As explained above, we uphold the disallowance in part and reverse it
in
part. For the "Stipends paid to ineligible entrants," we reverse
the
disallowance, except for those individuals who had not exhausted
their
18-month eligibility for CHEP assistance. For the disallowances
based
on "Daily classroom time and attendance records," we uphold ORR's use
of
a statistical sample here and, as explained in an appendix to
the
decision, we uphold ORR's determinations with regard to 9
contested
participants and reverse with regard to 17 participants. For
one
participant, we uphold the disallowance unless the State submits to
ORR
certain information within 30 days of receipt of the decision.
With
respect to cash contributions made by the Catholic Community
Services,
we conclude that ORR should make any adjustments to the universe so
that
it includes only stipend payments charged to CHEIAP funds. We
reverse
in full ORR's disallowance based on "Stipends paid for holidays" and
the
"Stipend reimbursement overclaim." We uphold ORR's disallowance
of
"Related indirect costs" to the extent the indirect costs are related
to
that part of the disallowance of direct costs we have upheld.
__________________________________
Donald F.
Garrett
__________________________________
Norval D. (John)
Settle
__________________________________
Judith A. Ballard
Presiding Board Member
. Appendix
In this appendix, we set out our individual determinations for
stipends
which ORR in its sampling process found were not adequately
supported by
time and attendance records. For a more general discussion
of the
issues involved, see the decision on pages 11-12. We address
the
stipends in the order used by the parties, identifying the
participants
by only their first names and the first initial of their last
names.
The dollar amount listed by each name is the amount of stipend at
issue
for each participant.
L., Julio -- $1.75
ORR should prevail; the State conceded that this amount was
properly
disallowed.
G., Jose -- $8.75
We find that the State should prevail. ORR questioned Jose's
attendance
in the program for five hours on December 29, 1982, apparently
because
Jose did not "sign out" at the end of that day, even though he
"signed
in" at 12:00 noon. We agree with the State that ORR has
not
demonstrated the inaccuracy of the payroll and attendance record
(PAR)
here, since we can assume that the counselor supervisor who prepared
the
PAR determined that Jose stayed until the end of the day at 5:00
p.m.,
absent evidence to the contrary. As State's Exhibit F-1
indicates, Jose
signed out for the previous day, December 28, at 5:00
p.m.
I., Maria -- $1.75
ORR should prevail; the State conceded this case.
M., Antonio -- $8.75
The State should prevail. ORR questioned this case because Antonio
was
paid holiday leave for Martin Luther King's birthday. As explained
in
our discussion of the holiday leave issue (see Decision, pp. 16-17),
we
conclude this was an allowable cost.
I., Celeste -- $43.75
ORR should prevail, unless, within 30 days from receipt of this
decision,
the State submits sufficient evidence to ORR that the $43.75
was not included
in the State's claim for this individual. The parties
in their
submissions agreed that Celeste never received a paycheck for
the
$43.75. ORR asserted, however, that the State had nonetheless
claimed
the $43.75 under its CHEIAP grant, and, therefore, the $43.75
should be
disallowed because no cost had been incurred. We agree with
ORR that
the disallowance was proper if the State claimed for a cost it
never
incurred.
The State did not have an opportunity to respond to ORR's assertion
that
the $43.75 was included in the State's claim, however.
Fairness
requires that the State should have this opportunity; this
finding
involves a substantial amount when projected to the universe of
stipend
claims.
F., Marie -- $38.50
ORR should prevail. The State explained that the auditors examined
a
"preliminary" time sheet which did not include Marie's name and
which
did not record attendance by any participants in the class on
June
29-30, 1983, the last two days of the pay period in question
(State's
Exhibit F-5). After the audit, the State submitted a new time
sheet
(State's Exhibit F-6), which recorded the attendance by all
participants
on June 29-30, and also added Marie's name.
We agree with ORR that the addition of Marie's name here
appears
questionable. State's Exhibit F-6 records clearly all the names
on the
earlier time sheet (albeit in a different order), but Marie's name
and
the hours for which she is noted are seriously marred, as if
other
notations were erased and written over. The State provided no
evidence
which supports its assertion that Exhibit F-5 was merely a
"preliminary"
time sheet or explains why an instructor would prepare a wholly
new time
sheet, rather than simply completing the existing one. We
conclude that
the questionable nature of Exhibit F-6 supports ORR's
determination to
disallow this expense.
A., Marie -- $8.75
ORR should prevail; the State conceded this case.
C., Carlos -- $8.75
ORR should prevail. The auditors found that Carlos was present in
class
40 hours for the pay period ending September 26, 1982. For the
next pay
period, ending October 10, 1982, Carlos was paid for an additional
five
hours which, according to the auditors, was intended to
compensate
Carlos for time spent in the program during the previous pay
period.
The State submitted a time and attendance record for the period
September
10 to 30, 1982, which records Carlos as being present a total
of 11 hours for
the period, including 2-3/4 hours on September 16.
State's Ex. F-7. ORR
argued that this exhibit did not explain the
additional five hours of pay,
since the auditors had earlier examined
documentation which showed Carlos to
be absent on September 16 and which
indicated that he had attended class for
only the 40 hours for which he
was initially paid.
We find that the State did not adequately support the additional
five
hours of pay to Carlos. Even if Exhibit F-7 is accepted
as
substantiating that Carlos attended class on September 16, it at
the
most shows a total number of 11 hours for the pay period, not the
40
that were claimed. Moreover, the attendance shown on Exhibit F-7
for
September 16 was for only 2-3/4 hours, not 5 hours, and we have
no
evidence connecting this attendance with the adjustment made in
the
later pay period.
G., Placido -- $35.00
ORR should prevail; the State conceded this case.
D., Nelson -- $8.31
The State should prevail. While ORR alleged that the time
and
attendance sheet was "altered" to add 4-3/4 additional hours for
one
day, an examination of the time sheets indicates that the 4-3/4
hours
was added to the time sheet to merely note the total hours attended
that
day. The original time sheet (State's Exhibit F-9) had a total of
"5"
hours for the other days of the week, but the total for Monday of
the
week in question was omitted, even though the starting time of 8:45
was
noted. The State reasonably explained that, when a participant
arrived
at the usual starting time of 8:30, only the total of "5" hours
was
noted; when a participant arrived at a different time, this other
time
was noted and the corresponding "total" hours were later filled
in. The
instructor here apparently failed to record the total for
Monday for
Nelson, which was later calculated by the counseling supervisor
who
prepared the PAR.
D., Ruossell -- $87.50
The State should prevail. The original attendance record was
apparently
lost, so the State instead submitted a notarized affidavit by
Ruossell
and the project director that Ruossell attended all classes during
the
pay period in question. See State's Ex. F-11. ORR's only
objection to
the affidavit was that it was
"suspicious" since the notary's
commission expiration date appeared to be in
1983, before the affidavit
was taken on November 30, 1984. However, we
do not find the affidavit
defective for the purposes here. The copies
of State's Exhibit F-11
submitted to the Board do not clearly indicate an
expiration date of
1983; although the stamped notary's form is blurred, our
copy of the
exhibit indicates the expiration date more likely to be "Jan. 31,
1985."
Moreover, the affidavit was witnessed by the project director and
we
have no other reason to suspect the authority or correctness of
the
affidavit. F., Raul -- $ .87
ORR should prevail. ORR explained that Raul was paid for 28.75
hours
for one week, but that he attended only 28.25 hours, according to
the
time and attendance records. See State's Ex.
F-12. The evidence
presented by the State does not refute this.
State's Exhibit F-13 is
some type of attendance record for Raul, with the
notations 2.25 and 2.5
for Friday, April 22, and with the letter A
(apparently for "absent")
for the rest of the week. This clearly does
not address ORR's point
that Raul was present for only 28.25 hours one week,
but received pay
for 28.75. State's Exhibit F-14, while hard to
decipher, also does not
address whether Raul was in class for 28.25 or 28.75
hours the week in
question.
C., Orlando -- $8.75
The State should prevail. ORR explained that the time and
attendance
records examined during the audit indicated that Orlando attended
class
for the pay period in question (June 20, 1983 - July 1, 1983) for
40
hours, but was paid for 45 hours. The State submitted an
attendance
record which indicated that Orlando actually attended class for 50
hours
during the pay period in question. State's Ex. F-16b. The
State argued
that Orlando was paid 45 hours for that week (State's Exhibit
F-17b),
but that he also received an "adjustment" of an additional five
hours
for the following week in order that he be fully compensated for the
50
hours in which he attended class. See State's Ex. F-18a.
ORR initially expressed doubts about the authenticity of State's
Exhibit
F-16b, which documented the 50 hours of class attendance.
Without
specifically addressing this exhibit's authenticity, we find that
the
other evidence submitted by the State supports the State's position
that
Orlando attended class for the 50 hours. The attendance record
attached
to the payroll register (State's Exhibit F-17a) substantiated the
50
hours attendance, and the same record for the next pay period
(State's
Exhibit F-18a) specifically recorded the positive adjustment of 5
hours.
Also, Exhibit F-18a was signed by Orlando, lending further
credibility
to the document.
P., Mirta -- $14.00
ORR should prevail; the State conceded this case.
R., Maria -- $3.94
The State should prevail. At the audit, the State presented
no
information to document an alleged discrepancy of 2.25 hours between
the
time and attendance records and the PAR for this participant for one
pay
period. After the audit, the State presented a document, which
ORR
described as documenting "time off" from the class. ORR questioned
this
document on the basis that it "appeared" that Maria's name had
been
added, but did not explain why it reached this conclusion. Without
some
evidence that Maria was not properly excused from class for the
2.25
hours (which might well not be recorded on the classroom time
and
attendance record the auditors originally examined), we find that
the
PAR provided sufficient support that Maria should have been
compensated
for the number of hours claimed by the State.
G., Evangelin -- $8.75
The State should prevail. The amount in question represented leave
for
holidays, which we have concluded in the decision was properly
allowed
by the State. See Decision, pp. 14-17.
M., Barbara -- $8.75
The State should prevail; same as above.
L., Francisco -- $8.75
The State should prevail; same as above.
S., Jose -- $8.75
The State should prevail; same as above.
C., Manuel -- $8.75
The State should prevail. The auditors noted that Manuel did not
sign
in or out on the time and attendance record for August 9, 1982,
although
he received a stipend for five hours for that day. After the
audit,
SFETC produced another time and attendance record, with "somebody
else's
initials" on the line where Manuel should have signed for August
9.
ORR's Post-hearing Memorandum, p. 4.
Without specific evidence that Manuel did not attend class or have
an
excused absence on August 9, we do not find it dispositive that
someone
else initialed the record that was submitted after the audit.
The PAR,
from which the amount paid was determined, presumably documented
that
Manuel either attended class or was excused on that day. The
initials
on the later form might have been those of the counselor supervisor
or
some other person who had authority to excuse Manuel for the day
in
question. We do not find that ORR here has impeached the correctness
of
the PAR.
V., Amorin -- $8.75
The State should prevail. The auditors here noted the fact
that
Amorin's time and attendance record was "corrected" to include the
time
for which she was paid but that she did not sign the appropriate line
on
the form to certify the correction. After the audit, the
subcontractor
located Amorin, who then signed the form to certify her
attendance for
the time in question. Even if ORR were on some ground to
question the
sufficiency of this later signature, ORR has not demonstrated to
us why
the time and attendance record (or the PAR) should be viewed
as
deficient merely because the participant did not sign the form
attesting
to the hours attended. ORR to our knowledge did not require
the
participant's signature on other time and attendance records
it
considered, and we do not view this case differently merely because
the
particular form here had a space for the participant to sign.
A., Georgina -- $3.50
The State should prevail. Although the time and
attendance sheet
examined by the auditors
did not reflect the time for
which
Georgina received $3.50 in compensation, the subcontractor
apparently
viewed the PAR as accurately reflecting the time in
question. State's
Ex. F-20. The subcontractor explained that an
investigation with the
counselor and participant indicated that the PAR
reflected a correction
to the attendance hours, which simply was not made to
the daily
timesheet until after the audit. ORR here submitted no reason
why we
should not view the PAR and the "corrected" time and attendance
record
as accurately reflecting time for which this participant deserved to
be
paid.
B., Marlene -- $17.50
The State should prevail. Amount represented payment for holiday
leave.
See Decision, pp. 14-17.
S., Jorge -- $8.75
The State should prevail. Although the time and attendance
record
examined by the auditors showed Jorge absent from
language-training
class on August 19, 1982, the State here submitted
substantial,
unrefuted evidence that Jorge was attending Employability Skills
and
Training and counseling on this day which was properly reflected on
the
PAR. Jorge signed a corrected time sheet to verify to his
attendance
(State's Exhibit F-22), the director of the subcontracting
agency
prepared a letter which confirmed his attendance for the day
(State's
Exhibit F-23), and the State submitted materials from the training
and
counseling session, as well as an evaluation form for that day
(State's
Exhibits F-24 and F-25). G., Carmelo -- $8.75
ORR should prevail. ORR noted that Carmelo was paid for five hours
on
November 5, 1982, but that the time and attendance record examined
by
the auditors indicated that Carmelo was absent on that day.
In
response, the State submitted an attendance record which
included
November 5, 1982 and which the State maintained demonstrated
Carmelo's
presence on that day. See State's Ex. F-26. ORR argued
that State's
Exhibit F-26 "does not show the additional five hours in
question."
The notations on State's Exhibit F-26 which purport to demonstrate
five
hours attendance on November 5, 1982 appear highly questionable to
us
and without further evidence we are unable to conclude that Carmelo
was
present on that day. The line on the form which appears to be
signed by
Carmelo is the only part of the form that is marred and hard to
read;
the participant's name on the left column of the form is
entirely
obliterated and all of the numerical entries appear to have been
erased
and possibly changed. The particular entry under November 5
is
impossible to read; while the notation could be intended as a "5,"
it
also could be any other number or none at all.
The State itself appeared to have questioned whether Carmelo was in
class
on November 5, 1982. The State submitted with regard to Carmelo,
"As
has been previously demonstrated, a participant may have been
actively
participating in a variety of ways; thus being absent from one
teacher's
classroom does not exclude him from being active in other
functions of the
program." State's Post-Conference Memorandum, p. 7.
However, the only
evidence submitted by the State purports to
demonstrate Carmelo's attendance
in class for that day. Without some
explanation of this discrepancy, or
clear evidence that Carmelo was
either in class or participating in another
function of the program, we
must agree with ORR that Carmelo's participation
on November 5, 1982 has
not been demonstrated.
A., Jose -- $8.75
The State should prevail. ORR disallowed this amount for the payment
of
holiday leave and specifically questioned holiday leave for
Christmas
Eve and for the Monday after Christmas when Christmas falls on
a
Saturday. State's Exhibit F-27 demonstrates that these days
were
treated as holidays by the subcontractor. (See our discussion on
pp.
14-17 of the decision.)
A., Lazaro -- $131.25
The State should prevail. ORR explained that the auditors found
no
indication in the time and attendance records that Lazaro attended
class
during the time for which he was paid. The State, however,
later
submitted documents to demonstrate Lazaro's attendance for the time
in
question, but which recorded the wrong social security number
for
Lazaro. ORR therefore questioned the "validity" of this
later
documentation. The State in response explained that the social
security
number recorded on this document must have been a mistake and
submitted
a computer listing of all program participants which showed that
there
was only one Lazaro A. in the program. State's Ex. F-28.
We agree with the State that the material which the
State later
submitted
provided sufficient corroboration of the PAR, notwithstanding
that the wrong
social security number was recorded on it. It seems
conceivable that
the social security number could indeed have been
recorded in error, and we
have no other indication that the material
submitted by the State was
unreliable. Further, the confusion in social
security numbers might
indeed be one explanation for why the auditors
thought the time and
attendance record initially examined did not
correspond with the
PAR.