DEPARTMENTAL GRANT APPEALS BOARD
Department of Health and Human Services
SUBJECT: Ohio Department of Human Services
Docket No. 86-162
Audit Control No. 05-60557
Decision No. 900
DATE: September 16, 1987
DECISION
The Ohio Department of Human Services (State) appealed a determination by the Office of Human Development Services (Agency), disallowing federal financial participation (FFP) in the amount of $1,356,032 1/ claimed under the Aid to Families with Dependent Children - Foster Care (AFDC-FC) program funded by Title IV-E of the Social Security Act (Act). Based on an audit for the period October 1, 1982 through March 31, 1984, the Agency found that the State improperly claimed FFP for institutional foster care costs not documented as aid or maintenance expenses. The Agency found that of the amounts claimed by the State, payments of $1,316,739 FFP to private, non-profit child care institutions (PNCCIs) and payments of $39,293 FFP to public child care institutions (PCCIs), for social services and other costs were unallowable under Title IV-E. In addition, the auditors found that of the $1,316,739 paid to PNCCIs and claimed under Title IV-E, $7,769 had also been claimed as social services under Title XX. 2/
Based on our analysis of the record, we conclude as follows:
o The disallowance of $39,293 related to PCCIs is upheld.
o The disallowance related to PNCCIs is upheld in the amount of $7,769 (see n. 2).
o The remainder of the disallowance related to PNCCIs is upheld provisionally, but, as discussed below, we remand this part of the disallowance to the Agency for further proceedings to give the State an opportunity to prove the amount should be reduced.
Background
Title IV-E of the Act provides federal matching funds to eligible states for "foster care and adoption assistance" to those children who would otherwise be eligible for assistance under Title IV-A, the Aid to Families with Dependent Children program. Title IV-E, however, limits the "foster care" part of the Title IV-E payments for purposes defined as "foster care maintenance payments" in section 475(4) of the Act. The reason for this limitation is that Title XX of the Act provides a separate funding mechanism for "social services" programs, which are specifically limited in amount. As reflected in regulations, one purpose of Title IV-E's definition of "foster care maintenance payments" was to exclude those "social services" covered by Title XX. See 45 CFR 1356.60(c)(3).
The Department of Health and Human Services Office of Inspector General conducted an audit of costs claimed by the State for institutional foster care of dependent children. The audit's stated purpose was to determine whether the State had administered the pass-through funds portion of the Title IV-E program in accordance with the Act and applicable regulations. 3/ Audit report, pp. 2-3. The auditors concluded that the reporting and control procedures established by the State for institutional pass-through costs were inadequate and failed to identify the unallowable social services and other costs included in the amounts reported by the County Departments of Human Services. Id., at p. 6. Since the records maintained by the State, the counties, and the institutions did not allow for precise identification of allowable maintenance costs, the auditors were obliged to devise their own methodology for calculating the amount of allowable costs to determine the amount to be disallowed.
In response to the draft audit report, the State conceded that its Title IV-E claim did not adequately distinguish between allowable and unallowable costs associated with social services in institutional settings. However, the State objected to the procedures used to compute the disallowance amount, and it submitted its own proposal for determining Title IV-E institutional maintenance costs, which would encompass all allowable maintenance costs for the audit period.
The Agency's regional office initially approved the State's proposal for calculating the allowable maintenance costs for the time period at issue. However, the Agency's Commissioner reject- ed the regional office's recommendation, ruling that the agreement violated section 475(4) of the Act and the regulations at 45 CFR 1356.60(c). 4/
Discussion
I. The Disallowance With Respect to PCCIs Is Correct.
The Agency was clearly justified in taking a disallowance here. The State conceded that it did not properly distinguish between allowable maintenance costs and unallowable social services costs in this case. State's Ex. A at p. 27. In addition, the State did not argue that the disallowance amount pertaining to PCCIs was incorrect. Therefore, we uphold the portion of the disallowance ($39,293) attributable to PCCIs.
II. The Amount of the Disallowance Related to PNCCIs Should Be Recalculated.
As noted above, the State admitted that its system did not adequately distinguish between allowable maintenance costs and unallowable social services costs. It was clearly the State's responsibility to do so, and thus the disallowance of PNCCI costs is certainly correct in principle. The State has, however, contested the Agency's calculation of the disallowance, arguing that the Agency has included in the disallowance amount some maintenance costs that are allowable. The first question, then, is whether the State has proved that the disallowance amount likely was overstated. We are mindful that the auditors were led by the State's own faulty claim to examine State records in an attempt to clarify the legitimate claim amount. Nevertheless, we find that the State has raised questions about the calculation amount sufficient to justify giving the State another chance to show an alternative amount, in accordance with the discussion below.
The kind of allowable costs at issue here are defined in section 475(4) of the Act, which provides:
The term "foster care maintenance payments" means payments to cover the cost of (and the cost of providing) food, clothing, shelter, daily supervision, school supplies, a child's personal incidentals, liability insurance with respect to a child, and reasonable travel to the child's home for visitation. In the case of institutional care, such term shall include the reasonable costs of administration and operation of such institution as are necessarily required to provide the items described in the preceding sentence.
The implementing regulations are found at 45 CFR 1356.60 and refer to the definition in the Act.
The State argued that the Agency's auditors violated the above-noted provisions by disallowing properly claimed costs. In the audit report, the auditors stated that to identify the maintenance portion of the reported payments to PNCCIs, they examined Title XX contracts, local agreements, institutional annual cost reports, and other documentation which supported the 115 negotiated rates billed by 35 institutions during the audit period. However, the auditors stated quite explicitly that they used only Title XX contracts, entered into by the county departments of human services, which were for "comprehensive services" (Audit report at 9; State's Ex. A at 13) for children who were wards of the county, to calculate the amount of the disallowance. Audit report, p. 10; State's Ex. A, p. 14. State procedures provided two types of services codes for institutional services at two different payment rates under these Title XX contracts. One service code provided for "an all inclusive rate for social services including counseling, therapy, and room and board, and a lower rate for social services including counseling, therapy, etc., but excluding room and board." Audit report, p. 10; State's Ex. A, p. 14 (emphasis added). After review of these contracts, the auditors concluded that the difference between the rates would represent the charge for foster care maintenance, which is the amount allowable under Title IV-E. Id.
The Agency did not deny the State's allegation, in its response to the audit report, that the auditors refused to recognize that social workers in the PNCCIs provided maintenance services, as well as counseling and therapy. Instead, the Agency argued that, since documentation that would allow precise identification of allowable maintenance costs was not available, the data used by the auditors was the best historical data available to them for the purpose of segregating maintenance costs from social services costs. The Agency maintained that the data used by the auditors is prima facie more reliable than the original data used by the State, when it first submitted its claim, because the State's data made no distinction between maintenance activities or rates and social services activities or rates. The Agency maintained that the Title XX contracts examined by the auditors and the rate breakdowns obtained from institutions did make such a distinction. Furthermore, as stated below, the Agency asserted that the auditors did not base their calculations of the unallowable payments to PNCCIs solely on this data. In any event, the Agency argued that the State should be estopped from challenging these procedures since it was the State's failure to maintain documentation which necessitated their employment. The Agency concluded that the disallowance must be upheld in its entirety unless the State can prove that the Agency erred in determining that the costs claimed were unallowable social services costs rather than foster care maintenance costs. 5/ We find that it is likely that the contract rates used by the auditors improperly excluded allowable maintenance costs. The contracts were for "comprehensive services" (Audit report at 9), which implies that more than strictly room, board, and social services per se were rendered. In addition, the Agency never contested the State's argument that documents examined by the auditors in addition to the contracts showed that costs incurred for social workers who performed maintenance activities were not included in room and board as defined in the contracts. Instead, the Agency erroneously claimed that those documents were taken into account (Agency response brief at 14) even though the auditors clearly used only the contracts in calculating the disallowance amounts (Audit report, p. 10; State's Ex. A, p. 14). Although the Agency never formally conceded in its briefs that social workers provided some maintenance services during the audit period, in the subsequent Random Moment Sampling (RMS) methodology adopted by the Agency's Regional Administrator for future claims, the Agency accepted that some allowable services were provided by social workers (State's Exs. C and H).
We find that the State has succeeded in raising questions about the Agency's calculation and general approach which are substantial enough to justify giving the State an additional opportunity to prove the amount of a fair reduction in the disallowance amount. We recognize that it is uncontested that the State's claim was seriously faulty from the beginning, and that the Agency should not be placed in the position of having to pay such a faulty claim merely because the Agency itself has had trouble in calculating precisely -- from State records -- the amount of unallowable costs. Thus, if the State fails to reasonably and timely establish the amount of its claim, the disallowance is upheld. The record here shows that the Agency, indeed, was not averse to allowing the State to recalculate its claim, if it could do so by means acceptable to the Agency. However, the Agency was unwilling to allow the State to use means which, in principle, have been generally accepted in other contexts (as we discuss below). Given our ruling below that good sampling techniques are an acceptable method for establishing the amount of the State's allowable claim, we urge the parties to attempt to resolve this calculation issue in the spirit of intergovernmental cooperation envisioned in the Act.
III. Although Sampling May Be Used to Establish Costs, the State's Proposed Methodology Is Faulty.
While we agree with the State that the Agency's calculation may have understated the amount of allowable maintenance costs, the State cannot shift its burden to document these costs as a condition for receiving FFP in them. As we said in Ohio Department of Human Services, Decision No. 858, April 8, 1987, and other prior Board decisions, a grantee has the burden of documenting the allowability of its claim. Further, it is a basic principle that in order to claim costs under a grant program, the grantee must show that the costs are necessary and reasonable to the grant program and are allocable to the program. Grantees are required to make and retain records of expenditures, and support these records with source documentation. 45 CFR Part 74, Subparts D and H.
In order to meet its responsibility for documenting claims, the State proposed a sampling methodology for calculating the amount of the maintenance expenses. The State proposed to calculate the disallowance through application of a RMS methodology developed subsequent to the audit period for the purpose of distinguishing maintenance costs from social services costs in future claims. RMS is a time study sampling procedure which consists of selecting, at random, a series of "moments" within a specified time frame and at each of these moments observing the activity of randomly selected workers in the child care institutions. The number of times the workers are engaged in the defined activity of interest is counted, and the count, divided by the number of moments, is taken as an estimate of the proportion of time that the sampled class of workers was engaged in the defined activity during the specified period. The State proposed that the percentages established under this methodology for the period July 1984 through June 1985 be applied to the State's cost reports for the audit period. 6/ In rejecting the State's proposed methodology, the Commissioner ruled that the methodology proposed by the State would be based on estimates, rather than actual expenditures, in violation of section 474(a)(1) of the Act and 45 CFR 74.61(b), (f) and (g), which implement that statutory provision. Section 474(a)(1) provides:
For each quarter beginning after September 30, 1980, each state which has a plan approved under this part (subject to the limitations imposed by subsection (b) of this section) shall be entitled to a payment equal to the sum of --
(1) an amount equal to the Federal medical assistance percentage (as defined in Section 1905(b) of this Act) of the total amount expended during such quarter as foster care maintenance payments under Section 472 for children in foster family homes or child-care institutions . . . .
In the proceedings before the Board, the Agency asserted that it rejected the State's proposal because the State's actual allowable costs for the audit period could not be calculated by use of this methodology. The Agency argued that the statute requires that claims for FFP may be submitted only for the actual expenditure amount. The Agency argued that sampling, at its best, produces only estimates of actual expenditures, and it maintained that, given its statutory and regulatory mandate to allow FFP only for the State's adequately documented actual expenditures, sampling could not be permitted.
Further, the Agency maintained that not only does sampling constitute an estimate of actual expenditures but the sampling method proposed by the State in this particular situation is also defective. The Agency asserted that, for cost claiming purposes, it can accept sampling only to the extent that it "is based on proper statistical procedures applied to actual experience during the period for which claims are to be made." Agency brief, p. 19; Agency supplemental appeal file, pp. 4-5. The Agency maintained that the proposed retroactive application of RMS does not meet this test because the historical data from the disallowance period, which is necessary to assure that the procedures are being applied to actual experience during the audit period, does not exist. The Agency submitted the declaration of the Chief of its Evaluation Branch (Gershenson declaration) to support its position. Agency's Ex. B. Further, the Agency maintained the State's data indicated that changes had occurred in the State's system in succeeding years, which rendered the proposed methodology unacceptable. Moreover, the Agency maintained, contrary to the State's assertion, that it has not approved the State's RMS methodology.
The State maintained that the proposed methodology is reasonable and an accurate method of calculating amounts of allowable costs. Moreover, the State argued that RMS is a method that has been proposed and accepted by HHS as such. The State argued that an Agency Policy Announcement, ACYF-PA-82-01, recommends RMS as an acceptable method for cost allocation. Further, the State maintained that it may apply sampling results retroactively if there is no significant difference in the program between the audit period and subsequent periods. The State presented an affidavit by its Chief of Division of Family Services (Schwerfager affidavit) to support its position that the level and type of care provided children in residential treatment/institutional care has remained substantially the same. State's Ex. O. Further, the State maintained, based on the Schwerfager affidavit, that throughout the years 1981-1985, the number of children in temporary and permanent custody has not varied more than 3.04%. The State also argued that it has adopted no regulations which have required changes in the type and characteristics of care received by children placed in residential or institutional care. Additionally, the State argued that if, as the Agency maintained, the State has provided more counseling and other unallowable social services in the succeeding years, this would result in an under-inclusive claim by the State because the State would be claiming less than actually allowable, since social services in the preceding years entailed less time. 7/ The Agency errs in its argument that sampling may not be used at all. As we discuss below, a valid sampling process is a well- established mechanism, and the statute simply does not preclude it. However, we agree with the Agency that the State's proposed methodology is seriously flawed.
It is well-established in Board and court precedent that a sound statistical sampling methodology can be used to determine the amount of costs properly charged to HHS programs. See California Department of Social Services, Decision No. 816, December 5, 1986, and decisions cited therein. As we said in California, sampling typically is used when a claim for federal funds arises where it is impossible, or at least costly and impractical, to examine each actual item of cost. Further, as we said in California and elsewhere, valid statistical sampling will produce a result which has a high degree of probability of being at least as good as a full-scale review. This is a basic reason why we cannot agree with the Agency's argument that the Act precludes the use of sampling. The Act contains no such specific preclusion, although we agree that the Act cannot reasonably be read to contemplate submission of mere loose "estimates" of claim amounts. 8/ A valid sampling process, particularly when applied in a way which places the risk of error on the proponent of the sampled data, produces a figure which is as good as, and may be better than, an "actual" figure subject to clerical errors.
Moreover, although the Agency maintained that it has not yet approved the State's sampling methodology, the Agency's policy announcement clearly suggests that sampling methodologies such as the one proposed by the State are permitted. Policy Announcement, ACYF-PA-82-01, at p. 6, states, in relevant part:
Various cost allocation methods, e.g., random moment studies or actual counts, may be used by institutions in developing their cost allocation plans. The State agency must approve the plan as a part of its approval of rates. (Emphasis added)
However, the question remains whether the State has shown that its proposed sampling methodology is valid. While sampling in its purest form envisions samples from the same period in question, common sense would dictate that samples from another period may be used if it can be established that no substantial change has occurred so as to invalidate the procedure. Cf. B.J. Mandel, Statistics for Management, p. 399, (1977). In the present case, the State has not provided sufficient evidence to support its position that its institutional foster care program has remained so constant that there are no significant differences between the data for the audit period and data for the subsequent period, which are necessary for an accurate calculation. Consequently, we cannot conclude that the State's proposed application of this sampling methodology to the audit period is valid.
The Agency's major objection to the use of the State's methodology is delineated in the Gershenson declaration submitted by the Agency, which noted that the State failed to provide statistical evidence of stability of the data between the audit period, October 1982 through March 1984, and the sample period, July 1984 through June 1986. 9/ The Gershenson declaration detailed the events which represented instability of the data as follows:
. . . In order to achieve compliance, and otherwise to improve its program operations, many changes in its Title IV-E foster care operation were necessary. For example, the May 1982 Program Review recommended 20 changes. . . . Moreover, during the entire period from 1981 through 1984 child abuse and neglect reporting increased from 16,514 in 1981 to 46,000 by 1984. The institutional population decreased from 2,960 in 1980 to about 1,800 by 1984. . . . Ohio's data concerning the ratio of child census to average Full Time Employee equivalents ("FTEs") from 1981 -1984 has a range from .82 to .98, a ratio difference of .16, which is, in fact, statistically significant. . . . The problem is compounded by the fact that the State has failed to show that even if the ratio of child census to FTEs was constant that this is equivalent or related to staff spending the same amount of time on all activities.
Although the State tried to explain away some of the inaccuracies in its data (Reply at p. 9), it never addressed the Agency's allegations that the data did not show consistency between the compared periods, nor did it respond to the Agency's contentions concerning the questionability of assuming constant level of care from the data presented. In particular, the State never denied the statement that the changed ratios of child census to FTEs within the audit period were statistically significant. Thus we agree with the Agency that the specific sampling methodology proposed by the State is unacceptable because the sample period and the audit period were not consistent.
In summary, we conclude that the Agency unreasonably rejected in principle the concept of use of sampling methodology to establish a reasonable disallowance amount. At the same time, the Agency reasonably rejected the State's proposed RMS methodology as flawed. For reasons already discussed, we believe the State ought to be given the opportunity to use a better methodology if it can and will do so. We urge the parties to cooperate in developing some approach acceptable to both. Common sense suggests that there must be a solution to this dilemma if the parties approach it in good faith; possibilities which occur to us are that the parties might revise or adjust the RMS method to offset potential errors, or develop a sounder data base for use with the RMS method, or agree to adjust the auditors' method to accommodate an agreed-upon factor for previously omitted allowable costs, or use a non-RMS methodology to extrapolate a figure from an intensive review of a sample of historical records.
Conclusion
Based on the analysis above, we conclude as follows:
l. The disallowance of $39,293 related to PCCIs is upheld.
2. The disallowance related to PNCCIs is upheld in the amount of $7,769.
3. The remainder of the disallowance related to PNCCIs is upheld provisionally, subject to Agency redetermination on remand in accordance with the following:
a. The State must propose in writing to OHDS a new or revised method for defining the amount of the disallowance, consistent with the discussion above, within 30 days after counsel for the State receives this decision (or within such longer period as OHDS allows). If the State fails to do so, the disallowance is upheld in full.
b. If OHDS rejects the State's new or revised method, it should notify the State promptly in writing, stating the reasons for the rejection. The State may return to this Board within 30 days after it receives this rejection. The sole issue for consideration will be the reasonableness of OHDS's rejection of the revised methodology.
c. We urge the State and OHDS to cooperate to try to find a reasonable, agreed-upon method for revising the disallowance amount, in accordance with the discussion above. We recognize that discussions during the proceedings before the Board were unsuccessful, but we anticipate that the determinations above will provide a basis for renewed cooperative efforts. To facilitate these. efforts, OHDS may modify the procedures in (a) above as it sees fit.
________________________________ Judith A.
Ballard
________________________________ Alexander
G. Teitz
________________________________ Norval D.
(John) Settle Presiding Board Member
1. The State did not appeal additional portions of the disallowance which upheld the auditors' findings.
2. Although the State argued to the Agency, in response to the audit report, that it had the option to claim allowable costs under Title XX or Title IV-E, the State did not attempt to advance its position before the Board or even comment on the Agency's finding that it made duplicate claims. In any event, in Ohio Department of Human Services, Decision No. 858, April 8, 1987, at pp. 10-11, the Board ruled that the State could not make duplicate claims under two federal programs because it would be contrary to the Cost Principles for State and Local Governments, Office of Management and Budget Circular No. A-87. We incorporate that discussion and finding here, and uphold this portion of the disallowance.
3. Pass-through costs are maintenance costs for children in foster care institutions which exceed the State's AFDC-FC standard need rate.
4. As a second basis for the disallowance, the Commissioner also ruled that the State was restricted from retroactively adjusting its claim back to October 1, 1982, because section 1132 of the Act and the regulations at 45 CFR 95.7 required that these claims be filed within two years after the calendar quarter of the expenditure. The Agency, however, notified the Board that it has subsequently decided not to assert these timely claims requirements as a ground for rejecting the State's proposal. See Agency brief, n. 2.
5. While not relevant within the context of this case, the Agency also argued that it specifically did not dispute social services costs for which sufficiently detailed information was not available, and for certain costs in counties not included in the examination. The Agency argued that this could have resulted in an additional disallowance estimated at $247,736 in FFP.
6. The State submitted that a two-part process would be used in applying the methodology. First, the State maintained that the annual cost reports submitted by each institution for 1981 through 1982 would be reconfigured to comply with the revised cost report that incorporates the institutional time study. Secondly, the new per diem rates would be used to recalculate the Title IV-E claims. State brief, p. 14; see also State's Ex. J.
7. The State apparently assumed that the child abuse and neglect statistics for the study period continued to show the rapid increase of the period 1981 - 1984, cited as support by the Agency for its conclusion that unallowable social services costs probably increased during the audit period. However, no such statistics were provided by either party, so that we cannot conclude, as the State argued, that the application of study period services time percentages to the earlier period would actually result in an underclaiming. The data do imply that the institutional child care picture was rapidly changing within the audit period.
8. We have, however, stated that under certain circumstances a valid claim may be based on an estimate derived from calculation or computation. See New York Department of Social Services, Decision No. 537, May 30, 1984; compare with New York Department of Social Services, Decision No. 542, June 4, 1984.
9. By motion dated August 5, 1987, the Agency moved to supplement the record with a letter by one of the State's Management Analysts to the American Public Welfare Association, which further substantiated the Agency's position that the State's data is inaccurate and unreliable. The Agency stated that it did not have a copy of the letter when it filed its earlier submission, and that it received the information as part of another matter. By submission dated August 12, 1987, the State objected to the admission of the letter and stated that the record should be closed to any additional submissions. Additionally, the State maintained that the letter added nothing new, and it cautioned the Board about the use of raw data, which can be misleading. Nevertheless, we find that the letter, at the very least, indicated that the State's raw data brought into question the accuracy of some of the State's records. The letter stated that the "CY '84 data represented only 80% unduplicated child counts." This statement indicated that at least some of the State's records are duplicative and inaccurate. While we do not rely on this letter for our decision, we find that the letter is relevant and we therefore admit it into the