DEPARTMENTAL GRANT APPEALS BOARD
Department of Health and Human Services
SUBJECT: Vanderbilt University Docket No. 86-140 Decision No. 903
DATE: October 6, 1987
DECISION
Vanderbilt University appealed the determination of the Regional Director, Region IV, upholding in major part a unilateral determination by the Division of Cost Allocation (Agency or DCA) of indirect cost rates for "on campus" organized research for fiscal year (FY) 1983 and FY 1984. The appeal involved disputes over the allocation to organized research of five areas of costs: library costs, departmental administration non-labor costs, professional liability insurance costs, Vanderbilt Professional Practice Program Business Office costs, and student services costs. We also considered arguments raised by Vanderbilt in this appeal that the Agency's determination for FY 1983 should be overturned because the Agency was bound by an indirect cost rate agreement for that year and that for both fiscal years the Agency and the Regional Director violated certain procedural regulations in their dealings with the University. The amount in dispute, itself a matter of disagreement, totalled $3,807,000 according to the Agency's last submission on the amount of disallowance. See DCA Summary of Final Determinations as of April 27, 1987.
For three of the disputed areas, library costs, departmental administration non-labor costs, and Vanderbilt Professional Practice Program Business Office costs, we uphold the Agency in full. For the costs of professional liability insurance and student services, we remand the issues so that the parties may determine a reasonable allocation, consistent with our analysis. We reject Vanderbilt's argument about the binding effect of a FY 1983 rate agreement, as well as its other procedural arguments.
One of Vanderbilt's concerns, articulated at the conference, was the Agency's failure to provide throughout the negotiation and appeal process the basis for its calculations for each of the disallowance amounts. The Presiding Board Member ruled that the Agency should provide this information, both for the amounts found unallowable at earlier stages and in terms of the Agency's present position before the Board. In response to this data, the University pointed out certain errors and other problems with the Agency's calculations, some of which were accepted by the Agency and served to reduce the disallowance for several cost areas. Tr., Vol. 7, p. 822. The Agency disputed several other proposed corrections by Vanderbilt; however, the parties did not complete their consideration of the calculation disputes during the Board proceedings. As the Presiding Board Member indicated at the conference, the parties may return to the Board if they are unable to resolve any remaining calculation questions in the implementation of this decision. Tr., Vol. 7, p. 830.
Below, we first explain the general background of the dispute and then analyze the issues. The sections of the decision analyzing the issues are as follows:
Section
I. Library Costs
II.Departmental Administration Non-Labor Costs
III. Professional Liability Insurance Costs
IV. Vanderbilt Professional Practice Program (VPPP) Business Office Costs
V. Student Services Costs
VI. The Finality of an FY 1983 Rate Agreement
VII. Vanderbilt's Concerns about Whether the Agency Acted in "Good Faith".
Page
5
23
40
51
60
72
83 Background: The Nature of the Dispute
Office of Management and Budget (OMB) Circular A-21 establishes the principles governing the recovery of indirect costs for federally funded projects at colleges and universities. The Circular is incorporated into Department regulations at 45 CFR 74.172. As explained in the Circular, indirect costs are costs which are incurred for "common or joint objectives" and which "cannot be identified readily and specifically with a particular . . . project or . . . activity." OMB Circular A-21, section E.1. An indirect cost rate is a ratio of total indirect costs to total direct costs; that ratio, expressed as a percentage, is applied to each project's direct costs to determine the indirect cost recovery.
A university's indirect cost rate is set based on a proposal submitted by the university analyzing costs for a particular year and proposing a rate using that year's costs. In proposing an indirect cost rate, a university will group its indirect costs by category and ultimately distribute each category's costs among its major functions. The indirect cost categories provided for in the Circular are: depreciation and use allowances, operation and maintenance, general administration and general expenses, departmental administration, sponsored projects administration, library, and student administration and services. OMB Circular A-21, section F. While there is some flexibility with regard to those activities designated as major functions, in general, such functions are instruction, organized research, other sponsored activities and other institutional activities. Section B. The proposed distribution of indirect costs to organized research at issue here involves three of the categories of indirect costs: library costs, student administration and services, and departmental administration (which includes the disputes over non-labor costs, professional liability insurance costs, and VPPP Business Office costs).
An indirect cost rate proposal is subject to negotiation between university officials and officials of the responsible federal agency. This Department, through the Division of Cost Allocation, is the responsible federal agency for approximately 95 percent of colleges and universities. Tr., Vol. 1, p. 20. If negotiations reach an impasse, as they did at Vanderbilt for the years under dispute, DCA may issue a unilateral determination setting an indirect cost rate. See Vanderbilt's Att. I.G. (DCA's unilateral determination of Vanderbilt's indirect cost rates for FY 1983 and FY 1984). If the institution disputes this determination, it may request reconsideration by the HHS Regional Director, who must provide the institution an opportunity to meet with him and to submit additional information. 45 CFR 75.6(b). Vanderbilt in this case availed itself of this opportunity. Vanderbilt's Atts. I.K.-I.O. Vanderbilt then exercised its right to appeal the Regional Director's decision to this Board under 45 CFR Part 16.
The appeal record indicates that the present dispute arose at a time of general concern within the federal government over rising levels of indirect costs at research institutions. Indirect cost levels have risen steadily at least since 1972. Agency's Ex. LLL (Report by the Comptroller General of the United States: Assuring Reasonableness of Rising Indirect Costs on NIH Research Grants--A Difficult Problem). Before the period under dispute here, the Agency had begun to examine more critically the indirect cost proposals of certain institutions. Agency's Ex. EEE (memorandum establishing DCA policy to perform on-site reviews at "high dollar" universities).
The issues presented here were hotly contested and have been extensively briefed. The parties introduced hundreds of exhibits and participated in an oral conference and hearing which extended over seven days. The issues themselves are inherently complex and concern a number of fundamental questions involving the interpretation and application of the Circular's provisions. Most of the issues have not previously been squarely addressed in this type of forum by either the Agency or the university community. The usual process of negotiating an institution's indirect cost rate, subject as it is to the give and take of compromise, would in most instances avoid the sharp clash of positions on the cost allocation issues presented here.
This is not to say, however, that we found ourselves struggling to apply administrative law standards to postures incidentally remaining after a failed negotiation. In each area discussed below, regardless of complexity, we were able to identify basic standards against which to test the reasonableness and persuasiveness of each party's evidence and argument. Our basic guidelines have been the principles of OMB Circular A-21, which we have analyzed according to the language of the Circular itself, general accounting principles, and other applicable precedent, as well as the underlying logic of the parties' positions.
I. Library Costs
OMB Circular A-21 provides a "standard" methodology for allocating library costs, but allows a university to devise alternatives in certain circumstances. For the years at issue in this case, Vanderbilt chose to use two alternative methods for allocating library costs, both of which used data concerning what Vanderbilt characterized as the "actual users" of the campus libraries. Vanderbilt argued that its use of this data provided a more equitable allocation of costs than the Circular's standard methodology, which Vanderbilt maintained was inferior because it was based on "potential," rather than actual, use of the libraries. The Regional Director rejected use of the University's proposed alternative methodologies for the years in dispute, finding that the standard base was the "more appropriate" approach. On this basis, the Agency disallowed $425,000 for FY 1983 and $423,000 for FY 1984.
Based on our analysis below, we conclude that Vanderbilt has a burden under OMB Circular A-21 to demonstrate that its alternative bases result in a more equitable allocation of library costs than the "standard" methodology, and that it has not met that burden. We therefore sustain the disallowance.
A. The Standard Methodology
The standard methodology, referred to as the FTE (Full Time Equivalent) base, involves a two-step allocation procedure for library costs of the University. See OMB Circular A-21, sections F.6.b. and c.; Agency's Brief, pp. 11-12 (explaining the methodology in full.) The FTE methodology first allocates the University's library costs to categories of users: students; "professional employees," including faculty members, staff, and students who are also employees; and a residual category of "other users." As the methodology's name implies, the allocation of costs among these categories takes into account the amount of time a student or professional employee spends in a particular role, on a full time equivalent basis. For instance, a graduate student who carries a course-load equivalent to that of a full time student and also works in a half-time position as a professional employee will be counted as 1 FTE student and 1/2 FTE professional employee.
Costs are further assigned in a "second level" allocation among the major functions of the University as follows. Costs assigned to students are allocated entirely to the instruction function. Costs assigned to "other users" are allocated to "other institutional activities," a miscellaneous category of costs which are not allocated to organized research. Costs assigned to the professional employee category are further allocated to the major functions of the university based on the wage and salary effort reports for all professional employees. See OMB Circular A-21, sections F.6.b. and c.
Section F.6.b. of OMB Circular A-21 provides that the FTE base is to be used "in the absence of the alternatives in Section E.2.d." Section E.2.d. provides in relevant part:
(1) Actual conditions must be taken into account in selecting the method or base to be used in distributing individual cost groupings.
* * *
(3) . . . [T]he distribution may be based on a cost analysis study which results in an equitable distribution of costs.
* * *
(4) If a cost analysis study is not performed, or if the study does not result in an equitable distribution of the costs, the distribution shall be made in accordance with the appropriate base cited in Section F [the standard FTE base for library costs], unless . . . the following condition[s] is met: (a) it can be demonstrated that the use of a different base would result in a more equitable allocation of the costs . . . .
Vanderbilt in this case did not perform a formal cost analysis study to support an alternative methodology,.nor did it choose to use the FTE base, but instead sought to demonstrate the advantages of an alternative methodology under section E.2.d.(4).
B. Vanderbilt's Alternative Approaches for Allocating Library Costs
Vanderbilt used two methods for distributing library costs among categories of users during FY 1983 and FY 1984. See Vanderbilt's Opening Brief, pp. 25-26; Vanderbilt's Conference Exs. 11-13. For the Medical School library, the University used a "registered cardholder base" which allocated the costs of that library based on the numbers of faculty, staff, graduate and professional students, and undergraduate students who had library cards. Cards were required to borrow materials from the library but were not required simply to enter the library for other purposes. For the other libraries for which some costs were allocated to organized research, the University used a "turnstile" base, which depended on a turnstile count of each of the user groups (faculty and staff, graduate and professional students, undergraduate students, other) who entered those libraries. The turnstiles worked by reading an identification card for each patron as he entered the library, and recording the population group to which he belonged. Tr., Vol. 1, p. 44.
After each library's costs were allocated according to the cardholder and turnstile bases, costs were further allocated in a similar fashion to that done by the standard FTE base (with some important differences, which we address below). Costs attributable to undergraduate students were allocated to instruction, and costs attributable to other users were allocated to other institutional activities. Costs attributable to faculty, staff, and graduate and professional students were allocated based on the University's salary and wage effort reporting system, which in effect measured the relative amounts of work effort devoted to each major function by the total number of individuals in each population group on campus.
Vanderbilt argued that use of the cardholder and turnstile bases resulted in a more equitable allocation of costs than the FTE base on the ground that each was a measure of "actual conditions" at Vanderbilt as to the use of library resources. See Vanderbilt's Opening Brief, p. 29; Vanderbilt's Reply Brief, pp. 10-11; Vanderbilt's Conference Ex. 10. Vanderbilt argued that this accorded with the directive of OMB Circular A-21 that "[a]ctual conditions must be taken into account in selecting the method or base to be used. . . ." Section E.2.d.(1); see Vanderbilt's Conference Ex. 10.
The Agency admitted that the FTE base was indeed a measure of "potential use" since it did not directly measure the degree to which the resources of a particular library were used by each of the University's population groups. Agency's Brief, p. 12. The Agency noted, however, that neither were the cardholder and turnstile bases true measures of "actual use," since both depended on assumptions about the relationship between the actual use of library resources and either the holding of library cards or the number of passages through a turnstile. Agency's Brief, pp. 15-19.
The Agency explained that the FTE methodology achieves an equitable allocation of costs since it relies upon the logical assumption that the allocation of effort on a campus-wide, aggregate basis should indicate how the resources of a particular library would be used by each category of user. Moreover, the FTE base avoids the theoretical and practical flaws which the Agency observed with both the cardholder and turnstile systems. See Agency's Brief, pp. 11-12.
On the basis of the record presented in this appeal, we agree with the Agency. We acknowledge that Vanderbilt's methodologies may offer some theoretical advantage over the FTE base in their measurement of "actual use," inasmuch as they use data pertaining directly to those persons who have either entered or are likely to use the libraries. However, we agree with the Agency that, in application, each of Vanderbilt's methodologies had major flaws. Because of these flaws, which we outline further below, we conclude that Vanderbilt has not met its burden to demonstrate that its proposed methodologies offered a more equitable allocation of costs than the FTE base. C. Vanderbilt's Burden of Proof under OMB Circular A-21
An important consideration in reaching our conclusion is Vanderbilt's burden under OMB Circular A-21 to justify its proposed methodologies. The Circular offers as a standard methodology an approach that would appear to work well in the aggregate, since it assumes in effect that the degree of use of a particular library by the different groups on campus will roughly parallel the allocation of resources to each of the institution's major functions campus-wide. Vanderbilt did not criticize this assumption or argue why it might not work fairly at Vanderbilt.
If a university decides to use an approach other than the standard methodology, the Circular first explains that one alternative for doing this is through a formal cost analysis study. If no such study is performed, then the institution should use the standard methodology, or finally, another methodology may be used, if a college or university can demonstrate that it is more equitable than the standard FTE base. See section E.2.d.(3) and (4). Section E.2.d.(3) further establishes strict requirements when a university uses a cost analysis study. The study must take certain statistical factors into account and be "statistically sound;" it must be sufficiently documented to allow federal review and must also be reviewed periodically and updated if necessary. These requirements reflect the federal government's obvious interest in ensuring the validity and reliability of an institution's proposed alternative methodology.
If an institution does not perform a cost analysis study, but elects to use an alternative approach, the Circular is silent on how the institution must specifically "demonstrate" that the methodology is "more equitable." However, given the strict requirements imposed on a university when it elects to use the approach of a cost analysis study, we find that the lack of any study available for review by the federal government here would make it crucial that the institution's proposal be fully and clearly supported. Thus, it is clear that the Circular imposed on Vanderbilt the affirmative and reasonable burden of proving, not merely that it used some rational allocation methodology, but, further, that its methodology was more equitable than the standard methodology. D. Analysis of the Medical School Library Cardholder Base
Vanderbilt supported use of the Medical School library cardholder methodology on the apparent basis that this was the best available measure of actual use during the period at issue. Vanderbilt's Opening Brief, pp. 29-30; Reply Brief, p. 14. Vanderbilt appeared to acknowledge, at least tacitly, that the turnstile base that was utilized at the other four libraries had some advantage over the cardholder base, and the University established that a turnstile system has since been implemented at the Medical School library. See Vanderbilt's Reply Brief, p. 14; Tr., Vol. 1, p. 44.
Unlike the turnstile base, the cardholder methodology did not represent current actual visits to the library, but instead looked at a pool of all individuals who might at some time use the library to borrow materials. Vanderbilt's Revised Conference Ex. 37. We find that the use of the Medical School cardholder base as a measure of "actual use" had two major flaws.
1. Cards may have been issued to users several years before the period being studied, and could have been issued as early as 1976. A listing of non-Medical School faculty who were cardholders indicates that many of the cardholders received cards in 1978 or other years well before 1983. Vanderbilt's Att. II.H. Vanderbilt presented no evidence that those issued a card during these earlier periods in fact used library materials during the time in dispute.
The only way in which the Medical School library cardholder list was updated was to account for those cardholders who had left the University. This was done on an annual basis. Vanderbilt's Reply Brief, p. 15.
2. Since a library card was required only to borrow materials, the cardholder base did not reflect other types of use of library resources, such as requests for information or use of the library building as a place of study. Indeed, there was considerable confusion during the appeal on the part of the University as to what the holding of a library card did indicate. The written briefs by the University indicated that cards were needed for borrowing purposes, but at the oral conference the University corrected itself to say that cards were necessary for entrance to the library. After checking further, upon inquiries from the Agency, the University again corrected itself to indicate that during the period in dispute cards were needed only to borrow materials. Vanderbilt's April 10, 1987 letter to the Agency.
Vanderbilt did present an affidavit and other evidence during the oral proceeding with the purpose of demonstrating that the cardholder base was a reliable measure of the use of library resources. Vanderbilt's Revised Conference Ex. 37. The affidavit, by the Director of the Medical School library, concluded that, based on the Director's "firsthand observation and experience," as well as that of his staff, the registered cardholder records for FY 1983 and FY 1984 were a "fair and reasonable approximation of the relative use of the Medical Center Library by the several user groups." Id.
We find that the director's affidavit is not sufficiently persuasive to change our analysis. His perceptions are conclusory generalizations, and he is an interested party. Also, his observations about the fairness of the "approximation" of use of library facilities begs the question "as compared to what," and we have no basis to attribute to him any particular knowledge of cost allocation issues or whether the cardholder base is a proper tool for measuring the allocation of costs. Further, we note that although Board rules and practice allow introducing into the record an affidavit prepared during this appeal, we are cognizant that opposing counsel and the Board did not have an opportunity to question the affiant about the basis for his conclusions.
Attached to the Director's affidavit was a "survey" of use of the Medical Library's photocopy service during 1983. Vanderbilt's Ex. D attached to Revised Conference Ex. 37. The survey was presented, according to the Director, to "estimate . . . the relative use of the journal collection by the several user categories." However, Vanderbilt did not explain what significance to derive from this survey, including, for instance, how the use of the journal collection might provide an estimate of the use of all library resources.
Vanderbilt also presented a comparison of its proposed allocation of costs for FY 1983 and FY 1984 with the preliminary results of a cost analysis study of FY 1986 data. See Ex. C to Vanderbilt's Revised Conference Ex. 37. The exhibit indicates some clear correlation between the results of the study and the proposed allocation for the earlier years. The Agency objected to the admission of this evidence, on the ground that the cost analysis study was entirely preliminary and had not been reviewed by the Agency. Tr., Vol. 5, pp. 426-429; see Agency's Brief, p. 6. While we admitted this material into the record over the Agency's objection, Tr., Vol. 5, pp. 442-443, we agree with the Agency that the significance of the evidence is diminished because of the tentative nature of the study's conclusions. OMB Circular A-21 specifically provides that a cost analysis study be subject to review by the Agency. Here, this study has not been subject to such review and the record does not contain sufficient information for us to fully evaluate the study. We also note there is a question whether it would even be appropriate to compare the results of a study of 1986 data to that concerning 1983 and 1984.
In sum, while we agree with Vanderbilt that as a theoretical matter a cardholder base might offer some measure of "actual use" that the FTE methodology does not offer, we also find that Vanderbilt has not demonstrated the overall reliability of its particular approach. In addition, as we consider below, there were a series of other problems found by the Agency regarding the implementation of both the cardholder and turnstile bases. While we do not accept all of the Agency's criticisms, we find in them ample support for the Agency's conclusion that use of the FTE base was more appropriate than the Medical School library cardholder base during the time in dispute.
Given this conclusion about the Medical School library cardholder methodology, we do not specifically address, as an independent matter, whether Vanderbilt's turnstile system alone could be properly rejected by the Agency. The parties in their discussions assumed the two methodologies to be inextricably intertwined and Vanderbilt did not argue that its two alternative bases were severable. Moreover, Vanderbilt in another context maintained that an FTE base could not be applied library by library, because Vanderbilt's accounting system could provide information on the full-time and part-time effort by graduate and professional students only on a "total library basis." Tr., Vol. 7, p. 900. Therefore, we do not have the option here of accepting the turnstile methodology and rejecting the cardholder methodology, even if we were to agree that in some circumstances a turnstile method might be more equitable than an FTE methodology, if applied correctly.
Given that we do not "sever" the use of the University's two methodologies, in our discussion below we consider of equal significance the criticisms of both the turnstile and cardholder bases. For instance, a criticism of only the cardholder system would be relevant to the overall issue of whether the Agency should accept use of both alternative methodologies.
E. The Agency's Criticisms of Implementation of the Two Methodologies
Besides issues already discussed above, the Agency's disallowance relied upon a number of other problems with the use of both the cardholder and turnstile bases. The Regional Director adjusted downward the amount that should be allocated to organized research because of three of these criticisms. See Regional Director's letter, pp. 3-5. For both the turnstile and cardholder bases, the Regional Director found that these adjustments resulted in a similar amount allocated to organized research as would the application of the FTE methodology. Regional Director's letter, p. 4. As explained below, .we accept some of these criticisms of the University's methodologies and reject others. The effect of our acceptance of some of the major criticisms is to support our conclusion to uphold the Agency's determination that the FTE methodology is more appropriate than the two proposed methodologies for the years in dispute.
1. The treatment of graduate and professional students
The Agency's most fundamental criticism of both the cardholder and turnstile methodologies was the treatment of graduate and professional students in the "second-level" allocation of costs. Agency's Brief, pp. 25-27. As we explained above, the University in its proposed methodologies would first allocate costs to the various user groups based on data obtained from the turnstile or cardholder systems. For those costs allocated to the categories of faculty and staff and graduate and professional students, the costs were then further allocated based on the salaries and wages of all the "primary users" of the libraries, i.e., those users who taught at, were enrolled in, or otherwise were directly associated with the school which the library in question was intended to primarily serve.
The University in the second level allocation of costs treated graduate and professional students in a similar manner as the University allocated the costs associated with faculty, examining only their salaries and wages for their time spent teaching or conducting research for which they were paid. See Vanderbilt's Reply Brief, pp. 26-27. The Agency argued strongly that this created a large distortion: the University in its calculations failed to account for the role of graduate and professional students as students, rather than employees. Agency's Brief, p. 25. While a graduate or professional student might devote some of his time toward activities for which he is paid, he obviously also spends substantial time receiving instruction. Vanderbilt's failure to account for this, argued the Agency, created an overallocation to organized research and an underallocation to instruction.
We agree with the Agency concerning the treatment of graduate and professional students. The Agency explained through a series of hypothetical examples how the University's allocation method here distorted the final allocation. See Agency's Exs. JJ2-JJ3. If the full amount of a graduate or professional student's time is considered to be for employment-related activities, clearly too many costs would be assigned to that function. Vanderbilt's approach here appears to reflect a simple oversight in its failure to account for the time that graduate and professional students spend in their role as students. The University disputed that this was an oversight and explained that it believed the student role of graduate and professional students was properly accounted for under its approach in two ways, as part of the "second level" allocation of costs. See Vanderbilt's Reply Brief, pp. 25-27; Tr., Vol. 1, pp. 53-56. First, some graduate and professional students work as teaching assistants, so their salaries and wages devoted to this activity would be included as part of the instruction function. Secondly, the instruction role of graduate and professional students was accounted for by the instruction activity of the faculty who taught those students. This activity by faculty was accounted for by the salaries received in connection with the faculty's teaching duties. Tr., Vol. 1, pp. 53-56.
This explanation by the University was not in our view responsive to the Agency's concern about the treatment of graduate and professional students. First, the time spent by graduate and professional students as teaching assistants is clearly independent of the time spent by these and other graduate and professional students in class and studying. Second, costs relating to faculty must themselves be allocated partly to instruction because of the time spent by faculty on their instruction duties. Again, the allocation of costs associated with faculty is entirely independent of the issue of the allocation of library costs associated with graduate and professional students, which clearly must also be distributed in a way which reflects their time spent using the libraries as students. Otherwise, the distribution fails to recognize the substantial time which graduate and professional students undoubtedly use the library in their capacity as students.
2. The exclusion of "missing" cardholders
Another criticism of the Agency, regarding the application of Vanderbilt's Medical Library cardholder methodology, was the exclusion of certain "missing" cardholders. Agency's Brief, pp. 23-24. After the University had initiated its appeal to the Board, Vanderbilt disclosed that a certain number of library cardholders had been omitted from both years' cardholder base. Agency's Exs. L, M. (The parties referred to these missing cardholders as "other users.") The number of cardholders omitted was quite significant; the Agency noted that for FY 1984 the omission had the effect of understating the number of cardholders by 30 percent. Agency's Brief, p. 24. The reason for Vanderbilt initially excluding these individuals, although not altogether clear from the record, was apparently because these users were from outside the Medical School; the University defined the other users as "hospital staff, Nashville Academy of Medicine Users, alumni and others." Agency's Ex. M.
Once the exclusion of these "missing" cardholders was discovered and considered, the University nonetheless maintained that these cardholders should not be included in the allocation of Medical School library costs. Tr., Vol. 1, pp. 168-169. The rationale offered by the University was that these individuals were minimal users of the library and that it would thus be unfair to consider them for purposes of allocating costs. Id. The University admitted it had no "statistical information" to support the conclusion that the use was minimal, but it nonetheless assured the Board that this was the case in practice. Id.
The Agency, on the other hand, objected to the exclusion of these cardholders and argued that the appropriate classification of this category of cardholders, under OMB Circular A-21, was as "other users," the costs for which are allocated to "other institutional activities." Tr., Vol. 1, p. 167. None of the costs in the "other institutional activities" category are allocated to federally funded organized research. The Agency emphasized that Vanderbilt had offered no evidence of the degree of use by these individuals. Tr., Vol. 1, p. 168. We also note that, even for those cardholders whom the University does count for purposes of allocating costs, use of the library might be minimal, since many of the cardholders received cards several years before the years in dispute. See our earlier discussion in Part D.
We find that Vanderbilt's inability to account for the significance of the use of the Medical School library by these "other users" illustrates one more practical problem in use of the cardholder base. If Vanderbilt had used some methodology which measured the degree of use by all the patrons of the library, this issue of how to classify these "other users" would not arise. As noted before, the number of cardholders who comprised this "other users" category was quite significant. The University clearly needed to provide by some means a reliable measurement of the degree of use by all user groups. Statements by counsel that use by certain individuals was minimal, with absolutely no supporting studies or material, is a clearly unacceptable way to do this. In the absence of some further specific support that the use by these "missing cardholders" was minimal, we would have to agree with the Agency that these "missing" cardholders should be classified as "other users" rather than merely excluded.
3. The issue of secondary clientele at the four "specialty libraries"
Another issue which illustrates the flaws in Vanderbilt's use of the cardholder base, as well as the turnstile base, is the treatment of "secondary clientele" at the four "specialty" libraries (the Medical School, Management, Divinity, and Education/Music libraries). See Regional Director's letter, p. 4, n. 2; Tr., Vol. 1, p. 173. These users included all those from outside the school which the particular library was intended primarily to serve, e.g., students from other disciplines who used the Medical School library. Vanderbilt's Opening Brief, p. 44. For both the cardholder and turnstile methodologies, these secondary clientele were counted in the "first level" allocation of costs on the basis of the relative use of the library by each of the population groups on campus. However, in the "second level" allocation of costs based on the salary and wage effort reporting system, the University's methodology only considered the salaries and wages of "primary users," i.e., those faculty, staff, and employed students from the school in question. Vanderbilt's December 29, 1986 letter to Board, p. 3.
For this reason, the Agency argued that this approach reflected an inconsistent treatment of the costs to be distributed and that it overallocated costs to organized research. Agency's Brief, pp. 22-23. The reason for this overallocation was that the Medical School library was the most significant library in terms of use of organized research dollars, so that by allocating the costs relating to all users on the basis of only the salaries and wages of the Medical School population, the methodology would be skewed by the higher proportion of salaries and wages to organized research by Medical School personnel, in relation to all other groups on campus. Tr., Vol. 1, p. 174. The Agency noted that for the Medical library, roughly 25 percent of users in FY 1984 were secondary clientele. Regional Director's letter, p. 4.The University had two responses to this criticism. First, the University explained that the data was not readily available to allocate costs in the second level distribution based on the allocation of effort by all users of the library. Vanderbilt's December 29, 1986 letter to Board, p. 4; Vanderbilt's Reply Brief, p. 21. As explained above, the cardholder and turnstile systems were designed to measure the number of library users and whether the user was faculty, staff, or student. The methodologies were not designed to identify the particular individual who used the library, with the purpose of allocating costs based on that individual's salary and wage distribution. (For the turnstile count, there was of course no record at all of the identity of a particular user; the turnstile system only detected if the user was faculty, staff, or a graduate/professional or undergraduate student.)
The University's other response to the Agency's criticism was that the treatment of secondary clientele was the same for all the libraries, so any alleged unfairness in the allocation of costs concerning the Medical School library would presumably be balanced by the treatment at the other libraries. Tr., Vol. 1, p. 176; Vol. 7, pp. 836-837. In other words, if costs associated with Medical School secondary clientele were allegedly overallocated to organized research based on the heavily organized research-oriented salary and wage distribution by Medical School faculty, the costs relating to secondary clientele at the other libraries would be underallocated to organized research based on the smaller devotion of effort to organized research by the primary clientele at these other schools.
In support of its approach to treating secondary clientele, Vanderbilt "surveyed" a group of secondary clientele to examine the degree of effort devoted to organized research. For 96 out of 171 secondary clientele at the Medical School library, Vanderbilt found that 20.78 percent of their effort was devoted to organized research. Vanderbilt's Opening Brief, p. 45; Vanderbilt's Att. II.H. (These 96 individuals were those still holding library cards at the time of the survey and were thus the only persons for whom Vanderbilt could access effort reporting system information.) While the Agency noted that this data was not selected on a random basis, Tr., Vol. 1, p. 177, we agree with Vanderbilt that this data may illustrate the unfairness of the Agency's approach to treat all secondary clientele as other users. However, we also agree with the Agency that Vanderbilt's approach of allocating the costs associated with the secondary clientele on the basis of salaries and wages of the primary clientele may be inconsistent and result in the unfair allocation of costs. Since the majority of library costs allocated to organized research are associated with the Medical School library, there may indeed be some total overallocation of costs to organized research. Medical School faculty, staff, and students presumably would have some higher concentration of effort to organized research than would the secondary clientele of their library, given the higher level of organized research in the Medical School than elsewhere. While a similar phenomenon may exist at the other University libraries, the greater significance of the organized research function at the Medical School relative to the other schools would appear to outweigh any compensating effect.
Without ultimately resolving the issue of how to allocate costs associated with secondary clientele, we find that the Agency's criticism is valid and lends further support to our conclusion regarding the inadequacies of Vanderbilt's proposed methodologies. The use of an FTE methodology would suffer none of these conceptual problems relating to secondary clientele.
4. The second-level allocation of costs at the Central/Science library
The Agency's final criticism of Vanderbilt's methodologies was the second-level allocation of costs at the Central/Science library. Regional Director's letter, p. 5, n. 3. The Central/Science library is, according to Vanderbilt, the campus' "central" library, Tr., Vol. 1, p. 186, but was still intended primarily to serve only the School of Arts and Sciences, according to the Agency. Agency's Brief, pp. 29-30. Vanderbilt in effect treated all users of the Central/Science library as primary users, allocating the costs associated with the users of the library (as measured by the turnstile system) on the basis of the salary and wage "effort distribution" for the entire campus, without considering the degree of use of the library by any particular group on campus. See Vanderbilt's Opening Brief, pp. 46-47. The Agency objected to this approach, arguing that the proper treatment of the users of the Central/Science library was to create a special "residual" category of persons who did not have their own specialty libraries (from the Schools of Engineering and Arts and Sciences), and, in the second-level allocation of costs, to allocate the costs associated with users of the Central/Science library on the basis of these individuals' salaries and wages. Regional Director's letter, p. 5, n. 3.
The Agency's rationale for creating this category of "residual users" was that use of the Central/Science library by those who have their own specialty library would be "minimal," in comparison to use by residual users. Id. Furthermore, since those with their own specialty library would tend to spend greater time on organized research than those in the School of Arts and Sciences, the University's treatment of the Central/Science library would otherwise create an overallocation of costs to organized research. Agency's Brief, pp. 29-30; Tr., Vol. 1, pp. 187-188.
Without ultimately deciding that one party's approach to allocating these costs was more reasonable, we find that the Agency's criticism of Vanderbilt's treatment of residual users illustrates another flaw with the turnstile methodology as implemented since, as we explain below, it recognizes that those with their own specialty library are likely to use the Central/Science library less often than those without their own specialty library.
One reason for Vanderbilt's criticism of the Agency's approach here of creating a category of "residual users" was that such an approach would in effect assume that those with their own specialty library make no use of the Central/Science library. This extreme assumption does, we find, appear unwarranted, based on Vanderbilt's own description of how its campus operated and Vanderbilt's explanation that the Central/Science library was in the center of campus and that the collection at the Central/Science library was supported by funding from schools with their own specialty libraries. Vanderbilt's Opening Brief, pp. 47-48; Reply Brief, pp. 28-29; Vanderbilt's Conference Ex. 1 (Campus Map).
However, even if Vanderbilt is correct that those individuals with their own specialty library make significant use of the Central/Science library, there is nothing in the record to establish that those with their own specialty library use the Central/Science library as often as the group of residual users. Obviously, if a medical student or a Medical School faculty member has his or her own library to use for research or study, he or she is likely to use the main Central/Science library less than someone associated with the School of Arts and Sciences, who has no specialized library to use.
In sum, while the Agency's approach of not considering any use of the Central/Science library by those with their own specialty libraries may appear extreme, we find that Vanderbilt's approach is flawed because of its failure to recognize that use of the Central/Science library by those with their own specialty libraries would likely be less than by the residual users.
Conclusion
In sum, given the burden on Vanderbilt under OMB Circular A-21 to demonstrate the superiority of its proposals, we find that the flaws in the basic Medical School library cardholder system, and in the implementation of both the cardholder and turnstile systems, made fully reasonable the Agency's determination to reject use of the two alternative methodologies in lieu of the standard FTE base.
II. Departmental Administration Non-Labor Costs
Another issue concerned how to allocate each academic department's general fund non-labor costs to the departmental administration (DA) cost pool. We first explain the terminology and relevant principles for distribution of non-labor costs. We then explain the .particular dispute here between the parties, which concerned the choice of a "direct charge equivalent" (DCE) formula. We finally present our analysis of the disputed DCE formulas and explain why we uphold the Agency on this issue. The Agency disallowed $249,000 for FY 1983 and $581,000 for FY 1984.
A. Background--Accounting for Departmental Administration Non-Labor Costs
The general funds used to operate each academic department at Vanderbilt consist of restricted and unrestricted funds. Restricted funds must be spent in accordance with the terms of the grant, contract, or other agreement making those funds available to the University. Unrestricted funds are used without such explicit requirements for their expenditure. Unrestricted funds are charged for non-labor costs (administrative and support services) that benefit each department's major functions. This dispute concerns how to determine the amount of non-labor costs paid from unrestricted funds to be included in the DA cost pool.
Non-labor costs directly benefitting organized research are charged directly and paid from restricted funds. However, as is typical with university accounting systems, the instruction function is not charged directly for non-labor costs.
Section F.4.a.(2)(b) of OMB Circular A-21 provides that administrative and supporting costs of an academic department "are allowable provided they are treated consistently in like circumstances." See section A.2.f. The Circular further provides in section F.4.b.(2) for the allocation of administrative expenses among the major functions of a department on a modified total direct cost (MTDC) basis. The parties agreed that a portion of the non-labor costs paid from unrestricted funds are properly included in the DA cost pool for distribution on a MTDC basis among a department's major functions. However, in order to abide by the requirement for consistent treatment of costs in like circumstances, it is necessary, since the instruction function is never directly charged for non-labor costs, to determine the amount of those costs paid from unrestricted funds that should be assigned to instruction as the Direct Charge Equivalent (DCE). The DCE is to be the equivalent, in nature and amount, to the non-labor costs directly charged to organized research and paid from restricted funds.
What is at issue here is the formula properly used to determine the DCE amount for instruction. The non-labor costs paid from unrestricted funds that remain after this DCE amount is deducted are allocated to the DA cost pool. Although a DCE formula is necessary to correct an inherent inconsistency in university accounting systems, there is no provision in OMB Circular A-21 specifying requirements for such a formula. See University of California Indirect Cost Rate, Decision No. 40, October 13, 1977, pp. 11-14 (explaining the use of a DCE formula in a different context). The general theory behind developing a formula, however, is that one can obtain the unknown DCE amount by assuming that it will bear the same relationship to a known category of costs as two other known categories bear to each other. In other words, the ratio of A compared to B (two knowns) equals the ratio of the DCE amount compared to another known, C. The ratio A/B may then be expressed as a percentage which is then multiplied by C (also referred to as the base amount) to obtain the DCE amount.
The allocation of non-labor costs to major functions of a department can be illustrated as follows:
General fund non-labor costs in a department � � � � _______________�_____________ � � � � Restricted fund Unrestricted accounts fund accounts � � � � � � Directly charged _____________�_______ to organized research � � � Application � of DCE � formula � � � � � Instruction � � � � DA cost pool Instruction/ � � \Other Sponsored Organized Other Activities Research Institutional Activities
See Agency's Ex. KK. .B. The Dispute--The Choice of a DCE Formula
The DCE formula proposed by Vanderbilt used the following ratio:
___Instructional salaries and wages__ Total restricted and unrestricted salaries and wages
Vanderbilt applied the resulting percentage to unrestricted fund account non-labor costs to obtain an amount which it allocated to instruction as the DCE amount.
Vanderbilt's approach assumes that the non-labor costs paid from unrestricted funds which should be "directly charged" to instruction (i.e., the DCE amount) bears the same relationship to total non-labor costs paid from unrestricted funds as instructional salaries and wages (which were paid only from unrestricted funds) bears to total salaries and wages paid from both restricted and unrestricted funds.
The Agency objected to this formula because of the inclusion of both restricted and unrestricted salaries and wages in the denominator of the ratio used. The Agency argued that, instead, only unrestricted salaries and wages should be included in the denominator. The Agency maintained that restricted salaries and wages were irrelevant here to the allocation of non-labor costs, since non-labor costs are directly charged to organized research and paid from restricted funds. The Agency viewed the calculation of this DCE as involving simply the allocation of unrestricted fund non-labor costs between two benefitting functions--DA and instruction--unlike Vanderbilt, which viewed the allocation as among three benefiting functions--DA, instruction, and organized .research. In other words, the Agency argued, the proper formula is the following:
Non-labor costs of instruction (unrestricted funds)
Non-labor costs (unrestricted funds)
equals:
Instructional salaries and wages __(unrestricted funds)Instructional and DA salaries and wages (unrestricted funds)
C. Analysis of the Parties' Positions on Vanderbilt's Proposed DCE
While Vanderbilt presented extensive materials in an attempt to support its DCE formula, we find, ultimately, that Vanderbilt's DCE depended on an illogical assumption that a ratio which used both restricted and unrestricted funds in the denominator could be used to allocate costs charged to purely unrestricted funds.
Vanderbilt in our view never adequately addressed this logical inconsistency. Instead, Vanderbilt appeared to rely primarily on the notion that there are a myriad of DCE's which could be used, and that the Agency had no basis on which to second-guess the particular formula chosen here by Vanderbilt. Vanderbilt did advance some conceptual criticism of the Agency's approach (addressed below), which it claimed supported its approach. The University, nevertheless, maintained that a DCE need not be totally logical in order to achieve a fair allocation of costs, and that Vanderbilt had provided sufficient support for its DCE. Tr., Vol. 2, p. 319. At the oral proceeding, Vanderbilt's consultant explained, in response to questions from the Board:
The idea is to get a reasonable amount associated with instruction prior to allocating DA. We can argue formulas and formulas, hundreds of different formulas, it all gets down to, is the formula that you're using at your institution going to give you an equitable amount to instruction prior to allocating DA on [an] MTDC basis.
PRESIDING BOARD MEMBER: Is it, in essence, what you are telling me is that these things don't have to work out logically or mathematically? CONSULTANT: That's right.
PRESIDING BOARD MEMBER: They just have to work out so that a number comes out -- and everybody agrees to it?
(Simultaneous conversation)
CONSULTANT: In a lot of cases, that is exactly what occurs. You have to have something that is proportional that has some reason to it. Okay? But the reason doesn't have to be mathematical precision.
Id. The Agency, however, asserted that an institution's DCE must indeed be inherently logical and mathematically sound.
As mentioned above, OMB Circular A-21 does not address the issue of the choice of a DCE formula, nor have we been presented with other guidelines specifically governing this. As with any matter involving cost allocation, the basic goal is the equitable distribution of costs. It is, practically speaking, impossible to obtain a result which will be the mathematically precise amount which would have been directly charged to instruction had Vanderbilt's accounting system in fact been set up to do this. Nevertheless, the theory behind the use of a DCE formula is to establish an equation which will give as accurate an estimate of an unknown amount (here, the DCE amount for instruction) as is possible under the circumstances. The formula, or equation, chosen must be based on a reasonable assumption that the unknown will bear the same relationship to one known category of costs which two other known categories of costs bear to each other. Further, once the formula has been established, the actual categories of costs used must correspond to the categories of costs set out in the formula. Thus, the competing DCE formulas can be evaluated by analyzing the soundness of any underlying assumptions and the actual application to known categories of costs.
The Agency's objection to Vanderbilt's proposed DCE may be shown by the following illustration of Vanderbilt's approach:
Labor Costs � Non-labor costs � _________________________ � ___________________________ �Instructional salaries � � �DCE for non-labor costs of� � and wages � � � instruction � �_ (unrestricted funds) � � � (unrestricted funds)____� � compared to: � compared to: ________________________ � _________________________ �Total salaries and � � �Non-labor costs � � wages (restricted and� � � (unrestricted funds) � � unrestricted funds) � � �_______________________� We agree with the Agency that Vanderbilt's DCE depends upon an illogical analogy between the distribution of labor costs and non-labor costs. The Agency had no problem with looking at the distribution of labor costs as a surrogate for non-labor costs, but stated that if an institution does this, it must compare that part of labor costs which are analogous to those non-labor costs being distributed. If one is distributing non-labor costs in only unrestricted fund accounts, one must examine labor costs charged only to unrestricted funds, and not also include costs charged to the restricted funds. In other words, the denominator of the ratio of the two known categories of costs must have the same characteristics as the other denominator in the equation. The Agency expressed this as a requirement for consistency between the distribution base and the pool of costs to be distributed. Tr., Vol. 6, p. 614.
The Agency explained and presented an affidavit to the effect that the formula advanced by the Agency (including only unrestricted salaries and wages in the denominator of the labor costs ratio) was one of the two DCE formulas which are commonly accepted by those in the field of university cost allocation. Tr., Vol. 2, pp. 328-329; Agency's Ex. TT. The Agency maintained that its proposed formula was the approach known as the University of Southern California (USC) methodology. The other commonly accepted DCE, according to the Agency, is the University of California (UC) methodology.
Vanderbilt presented its own understanding of the USC and UC DCE formulas, which the University argued were "supportive" of its proposed DCE, since those formulas resulted in a similar or greater allocation of costs to the instruction function. We address in a later section the merits of these and the other supporting DCE's presented by Vanderbilt. The Agency presented an affidavit prepared for this case by the acknowledged originator of the UC and the USC methodologies, who averred in effect that it was the Agency's, and not Vanderbilt's, definition of these formulas which were what he had initially proposed. Agency's Ex. TT. The affiant's statement here is of course not dispositive on the issue of which party's DCE should be used; nonetheless, we find that this affidavit does lend some authority for the soundness of the Agency's approach.
Vanderbilt described and attempted to document an imperfection in the Agency's proffered DCE, which Vanderbilt maintained lent support for use of the DCE formula which it instead proposed. The University alleged that the Agency "erroneously assumed that all non-labor costs that benefitted sponsored research were specifically identified and charged directly to sponsored projects." Vanderbilt's Opening Brief, p. 59. In other words, argued Vanderbilt, some significant amount of non-labor costs were not directly charged to organized research, but nonetheless directly benefitted organized research. Tr., Vol. 2, pp. 289-290. As a result, according to Vanderbilt, the application of a DCE formula that is intended just to counterbalance the direct charges to organized research would overallocate costs to instruction. We explain below why we find Vanderbilt's evidence insufficient to establish that any significant amount of costs that would be properly categorized as direct charges to organized research were charged to the unrestricted fund accounts and why, even if Vanderbilt had established this, we would find Vanderbilt's proposed DCE formula unacceptable. Thus, we reject Vanderbilt's criticism as a basis for supporting the University's position in this matter.
In order to demonstrate its allegation about the direct charging of non-labor costs, Vanderbilt presented, in response to requests for information from the Agency, a series of accounting schedules which listed certain types of non-labor expenses by "object codes." See Vanderbilt's Atts. III.B-III.M; Vanderbilt's Opening Brief, pp. 63-67. Vanderbilt listed those object codes for which zero percent of costs were directly charged to organized research, those for which less than 10 percent were directly charged to organized research, and those for which less than 25 percent were directly charged. Vanderbilt sought to prove by means of these exhibits that, if none of or a small amount of the costs incurred in these object codes were directly charged to organized research, then one would logically expect costs paid from unrestricted funds to benefit organized research and thus to be analogous to those directly charged to organized research and paid from restricted funds.
We conclude that the data presented by Vanderbilt does not demonstrate its point. The Agency observed that, at the Medical School, for which Vanderbilt performed a complete analysis of the charging patterns by object code, only three percent of the total non-labor costs fell into the category of zero percent directly charged to organized .research. Agency's Brief, p. 39. For those object codes in which less than ten percent or less than 25 percent was directly charged to organized research, the Agency properly noted that Vanderbilt presented no evidence to demonstrate that any costs in the remainder of each of these categories of non-labor costs directly benefitted organized research. Agency's Brief, pp. 40-41. The data alone does not demonstrate a pattern of misallocation between the restricted and unrestricted fund accounts for costs directly benefitting organized research.
While Vanderbilt relied on the alleged general lack of precision with which its accounting system could track costs incurred for organized research because of an inability to monitor all supplies and services used by research personnel, this is not sufficient to establish that any significant amount of costs which should have been direct charged to restricted funds were not. One of Vanderbilt's examples was that a researcher might take a pencil out of a supply cabinet to use on a sponsored project without charging its cost to the appropriate restricted fund account. Tr., Vol. 2, p. 286. While Vanderbilt provided no evidence that this did in fact occur, we recognize that it is a possibility. It seems just as likely, however that a researcher might inadvertently use a pencil direct charged to a sponsored project for instructional purposes. Another problem with Vanderbilt's argument is that, even if researchers fail to direct charge everything they could, the likelihood is that these items are relatively insignificant given the obvious incentive to charge costs to restricted funds. See Agency's Ex. TT, p. 2, para. 5. Indeed, Vanderbilt acknowledged that "most large ticket non-labor items" were direct charged. Tr., Vol. 6, p. 633. We also are unwilling to assume that Vanderbilt's accounting system was as flawed as their argument would imply.
In evaluating Vanderbilt's argument, we also looked at the distribution of total non-labor costs for a given year. At the Board's request, Vanderbilt presented two exhibits at the oral conference which summarized this data for 1984. See Vanderbilt's Conference Exs. 39 and 40. Out of total general fund non-labor costs for that year of about 13.5 million dollars, 51.66 percent were direct charged to the restricted fund accounts and 48.34 percent were charged to unrestricted fund accounts. As a matter of simple common sense, the amount of non-labor costs direct charged to organized research (restricted fund accounts) argues against Vanderbilt's allegation that a disproportionately small amount of non-labor costs was directly charged to organized research.
Vanderbilt's Exhibit 40 also indicated that the DCE amount determined by applying Vanderbilt's formula was slightly less than half of the total non-labor costs in the unrestricted fund accounts for 1984. In other words, less than half as many non-labor costs were "directly charged" to instruction as were directly charged to organized research activities. This also does not appear to correspond with Vanderbilt's view that non-labor costs directly charged to organized research were understated.
Furthermore, even assuming the truth of Vanderbilt's assertion that a significant amount of non-labor costs paid from unrestricted funds directly benefitted organized research, the University did not explain how the formula it sought to use would correct this. The only logical support we see in Vanderbilt's approach is that, since the Agency's DCE allegedly produces an overallocation to instruction, Vanderbilt's DCE would produce some lesser allocation. Vanderbilt has not demonstrated to us how the application of its DCE formula is designed to specifically cure the imperfection which it argued exists in the DCE applied by the Agency.
While we understand Vanderbilt's basic point to be that unrestricted non-labor costs also directly benefit organized research, so that the salaries and wages from restricted funds ought to be in the base of the DCE ratio, Vanderbilt's formula also implicitly assumes that the alleged benefit to organized research is proportional to the benefit to instruction without recognizing that a great many non-labor costs are, in fact, directly charged to organized research. Also, we see no logical basis for correcting this alleged problem by increasing the pool of costs to be distributed as indirect costs via the DA cost pool. Section E.2.b. of OMB Circular A-21 requires that a cost grouping "should constitute . . . a pool of those items . . . of like nature in terms of their relative contribution to (or degree of remoteness from) the particular cost objectives to which distribution is appropriate." Including costs directly benefitting organized research in the DA cost pool would violate the requirement that costs be of "like nature."
While this provides additional support for our decision, we find most persuasive here the Agency's point that there is no basis to find from the object code data that the non-labor costs that ought to have been charged directly, in fact, were not charged directly and paid from restricted funds. Our acceptance of this point establishes the correctness of the assumption underlying the Agency's DCE formula that the non-labor costs benefit instruction and departmental administration only.
Another reason offered by Vanderbilt in support of its DCE approach, as opposed to the Agency's, is that the Agency's formula would result in an unrealistically small amount of non-labor costs for inclusion in the DA cost pool. According to Vanderbilt, the Agency's DCE would produce a ratio close to one (i.e. one which would give a percentage close to 100 percent), since instructional salaries and wages (in the numerator of the DCE ratio) are such a large part of unrestricted salaries and wages (the denominator of the DCE ratio advocated by the Agency). Tr., Vol. 2, p. 282. Vanderbilt presented no salary and wage information to illustrate this point. However, assuming that the DCE formula advocated by the Agency did produce a high ratio, we fail to see why this would provide a logical basis for rejecting its use, absent evidence that the amount allocated to DA was in fact unrealistically small.
Vanderbilt did not contest the general assumption that the distribution of salaries and wages is a basically reliable surrogate for the distribution of non-labor costs. Indeed, this is the primary assumption of Vanderbilt's proposed DCE formula. Therefore, if a large proportion of salaries and wages are devoted to instruction, even for only the unrestricted fund accounts, one should conclude that a similarly large degree of non-labor costs are allocable to the instruction function. If the ratio produced by the DCE formula is close to unity, we see nothing unreasonable in concluding that close to all non-labor costs for the unrestricted fund accounts are indeed allocable only to instruction.
D. Vanderbilt's "Supporting" DCE's
In support of its DCE approach, Vanderbilt produced a series of three "supporting DCE's" which Vanderbilt maintained resulted in an allocation of costs to the instruction function similar to that produced by the proposed DCE. We briefly analyze below these alternative approaches, but emphasize that our primary rationale for our decision on the issue of DA non-labor costs is that we find Vanderbilt's primary approach to be based on a fundamentally illogical assumption. That some other DCE, even assuming it to be based upon another approach with some conceptual validity, might result in the allocation of a similar amount for a particular year would clearly not of itself be directly dispositive as to the logical foundations of the DCE actually proposed by the University.
Vanderbilt's first "supporting DCE," apparently devised for the specific purpose of "supporting" the University's proposed DCE that we discussed above, worked as follows. For all "pure instruction" departments (which performed no organized research activity), the University divided general fund non-labor costs by total general fund salaries and wages. The ratios from each department were then averaged, and the resulting ratio served as part of the DCE formula to allocate unrestricted fund non-labor costs for each research department. Vanderbilt's Conference Ex. 19.
We agree with the Agency that this "supporting DCE" is unreasonable, for the two reasons set forth by the Agency in its brief and oral argument. E.g., Agency's Brief, pp. 41-44. First, the formula seeks to compare total non-labor costs for certain departments with total salaries and wages. Yet, the resulting ratio was then applied by Vanderbilt to non-labor costs in each department, rather than to salaries and wages. This is inherently illogical, since the formula would be taking the ratio of non-labor costs to labor costs, yet using that comparison as a basis for making conclusions about the relationship of non-labor costs to other non-labor costs. As the Agency noted, the only proper use of the University's "supporting DCE" ratio might be to apply it to instructional salaries and wages in other departments. In other words, it is the Agency's point that it is illogical to equate the relationship between general fund non-labor costs and total general fund salaries and wages (for instruction only departments) to the relationship between the unknown non-labor costs (unrestricted funds) for instruction and total non-labor costs (unrestricted funds) (for departments performing organized research).
Second, we question whether Vanderbilt's supporting DCE is based upon a proper analogy in its examination of instruction-only departments. (This is also the premise of the first of the other two "contingent supporting DCE's" which we address below.) The basic premise of examining non-labor costs in instruction-only departments is clear: the degree of use of non-labor costs in those departments involving only instruction should roughly parallel the use of non-labor costs attributable to the instruction function in those departments where both instruction and organized research are performed.
However, the Agency observed a weakness with this analogy which Vanderbilt in our view did not meaningfully rebut. Regional Director's letter, p. 8. The degree of non-labor costs associated with instruction-only disciplines, such as English or Fine Arts, might be entirely different than the degree of non-labor costs incurred in connection with the instruction or teaching function of departments involving organized research. For instance, a chemistry professor might need unusually expensive equipment when teaching students, while an English professor might not require any significant use of equipment or materials at all.
In its initial brief to the Board, Vanderbilt asserted that, contrary to the Agency's analysis, the non-labor costs for at least one instruction-only department actually exceed that for an organized research-oriented department. Vanderbilt's Opening Brief, p. 73. To demonstrate this, Vanderbilt presented a calculation in which it apparently divided the amount of non-labor costs by the number of full-time equivalent (FTE) employees for an instruction-only department, English, and for an organized research department, Biology. The resulting ratios were $1,456 per FTE in Biology and $2,703 per FTE in English. Id., Vanderbilt's Att. III.Q.
The Agency disputed these figures and explained at the oral proceeding that the basis for the University's calculations was that Vanderbilt had examined only the non-labor costs in the unrestricted fund accounts for the Biology department, yet compared this amount to all personnel, whether or not their activities were involved with unrestricted or restricted funds (for which associated costs are directly charged). Once non-labor costs for the restricted accounts are added into the calculations, the comparable ratios are, in round numbers, $4,300 per FTE in Biology and $2,700 per FTE for the instruction-only department. Tr., Vol. 6, pp. 653-655; Agency's Ex. OO. The Agency also performed a series of other calculations to compare the extent of non-labor costs in the Biology, English, and Fine Arts departments, which confirmed the point that non-labor costs are indeed higher for the research-oriented Biology department. For example, the ratios of non-labor to labor costs in the unrestricted fund accounts for Biology, Fine Arts and English were, respectively, 16.8 percent, 7.3 percent, and 3.6 percent. Agency's Ex. OO.
Vanderbilt additionally presented two "contingent supporting DCE's," which the University alleged also confirmed the amount allocated to instruction by its proposed DCE formula. The first of these Vanderbilt labeled the "University of Southern California" DCE. See Vanderbilt's Conference Ex. 20. The Agency disputed that this was in fact the DCE formula that is generally accepted by that name. Regardless of the degree of general acceptance of this DCE formula, we agree with the Agency's analysis that this formula does not serve as a basis to support Vanderbilt's proposed DCE formula. As with the first "supporting DCE," this University of Southern California DCE examined instruction-only departments, dividing non-labor costs for these departments into total salaries and wages. The result was again averaged between these departments and this ratio was used to calculate DCE amount. However, unlike with the "supporting DCE," Vanderbilt then applied the resulting percentage to each department's salaries and wages, rather than to non-labor costs.
As Vanderbilt noted, this approach satisfied the Agency's concern that a ratio with a base including salaries and wages be applied only to other salaries and wages. However, this "USC" methodology still relied upon the basic assumption that the ratio in instruction-only departments of non-labor costs to labor costs would be the same as the ratio of non-labor costs to labor costs for the instruction function in research-oriented departments. As we found above, the Agency presented convincing data that this assumption is unwarranted. Vanderbilt's second contingent supporting DCE was labeled the University of California methodology (which, again, the Agency said was not the more commonly accepted approach by that name). This DCE worked as follows. On a campus-wide basis, the non-labor support costs charged directly to organized research (for the restricted fund accounts) were divided by salaries and wages for organized research. The resulting ratio was applied to unrestricted non-labor costs to calculate the DCE amount. Vanderbilt's Conference Ex. 21.
As the Agency also remarked in one of its criticisms of Vanderbilt's supporting DCE, this "UC" approach is based upon the highly questionable assumption that a ratio with a denominator comprised of salaries and wages could be applied to a pool of costs comprised of non-labor costs. As we found above in this part, this is totally illogical.
E. Conclusion
In sum, based on our analysis above and the record on this issue, we conclude that the DCE formula advocated by Vanderbilt is clearly unreasonable. We found Vanderbilt's formula to be based upon a fundamental inconsistency between the treatment of restricted and unrestricted accounts. While the Agency's proffered DCE formula may not be the only way in which to allocate non-labor costs, Vanderbilt did not demonstrate that the Agency's approach produced a significantly unfair allocation or that Vanderbilt's DCE formula operated to correct this.
III. Professional Liability Insurance Costs
Vanderbilt allocated the professional liability insurance costs for certain Medical School faculty through the DA cost pool. (See Section II for our explanation of how unrestricted fund non-labor costs are allocated among a department's major functions.) The Agency acknowledged that some part of these insurance costs are allocable to organized research, noting that "medical malpractice insurance costs at Vanderbilt benefit sponsored research." Agency's Brief, p. 48. However, the Agency objected to the manner in which Vanderbilt allocated these costs to the major functions of each department, applying the same modified total direct cost (MTDC) base as with other non-labor costs paid from unrestricted funds. According to the Agency, this produced an overallocation to organized research, since the risk of liability claims arises primarily from patient care activities. Because of the alleged inequity in including these costs in the DA cost pool, the Regional Director disallowed all liability insurance costs allocated to organized research, resulting in a total disallowance of $30,000 for FY 1983 and $139,000 for FY 1984.
We find that Vanderbilt's inclusion of these costs in the DA cost pool creates an overallocation to organized research, but we remand the matter to allow the parties to devise an acceptable approach for allocating these costs consistent with our analysis below.
The Agency in its brief proposed that these insurance costs be removed from the general fund non-labor costs and allocated to organized research using an alternative base the Agency had developed. The Agency presented this base "for consideration by the Board." Agency's Brief, p. 50. While, as noted above, we accept the Agency's general point that Vanderbilt's methodology produces an overallocation to organized research, we reject the Agency's proposed base. In our view this base relied on a position we find overly restrictive, i.e., that organized research projects, other than those involving human subjects, do not generate a risk of malpractice claims and derive no allocable benefit from professional liability insurance costs.
We first explain below Vanderbilt's system for determining the amount of liability insurance premiums assigned to a given department for allocation through the DA cost pool. We then address why we accept the Agency's argument that the use of an MTDC base creates an overallocation to organized research. We last explain why we have rejected the Agency's proposed formula.
A. Vanderbilt's Rating and Modifier System
The amount of liability insurance premiums Vanderbilt assigns to a department depends on the premium amount for each of its faculty physicians for whom the University pays professional liability insurance. The premium amount for a particular faculty member is determined through a rating and modifier system, as explained in Vanderbilt's brief. See Vanderbilt's Opening Brief, pp. 89-92. The rating for a particular physician is set according to his or her area of practice and is determined using a system of the Insurance Services Office (ISO), identified by Vanderbilt as an "advisory bureau which maintains statistics and develops forms used by most property and casualty insurance companies." Vanderbilt's Att. IV.I.
The rating for a physician is then "modified," or reduced, because of several factors, including involvement with research projects. According to Vanderbilt, the modifiers work as follows. The rate for "fellow" physicians is reduced by 25 percent, since fellows are considered to be in training to develop specialties. The rate for residents is reduced by 50 percent, in recognition of their participation in a formal training program. Rates are also modified by 50 percent if a physician's effort is devoted more than 75 percent to research, if a physician's effort is devoted more than 90 percent to administration, or if the physician has a practice in a Veterans' Administration hospital. Vanderbilt emphasized that the rating and modifier system also determines what amount of insurance premium costs are designated to the University Hospital, rather than to the Medical School departments, presumably based on some measure of each faculty member's hospital duties as compared to his or her department duties.
B. The Agency's Criticism of Vanderbilt's Use of an MTDC Base for Allocating Liability Insurance Premiums
The Agency's primary criticism of Vanderbilt's treatment of liability insurance premiums did not concern the rating and modifier system itself, but rather Vanderbilt's treatment of these costs like any other DA cost, so that insurance costs were allocated on an MTDC basis to all of the department's major functions. The Regional Director concluded, "[I]t appears that these insurance costs simply do not benefit the various cost objectives of the DA cost pool like the other types of non-labor costs." Regional Director's letter, p. 9.
The use of a modified total direct cost base in allocating the DA cost pool means that these costs are allocated to each of the institution's major functions depending on the relative distribution of salaries and wages and certain other costs associated with that function. OMB Circular A-21, section G.2. The Agency's objection to using this approach for allocating liability insurance costs was that this assumes in effect that each of a department's functions would play an equal role in generating the need for liability insurance. The Agency observed that this assumption was unwarranted since, logically, patient care activities would be primarily responsible for the need for liability insurance. This phenomenon is, according to the Agency, recognized under Vanderbilt's own rating and modifier system. Under Vanderbilt's modifier system, the lesser significance of research in the generation of insurance costs is indicated by the 50 percent "discount" applied to a faculty physician's rating if he or she is involved more than 75 percent of the time with research. See Tr., Vol. 3, pp. 509, 524. We consider it self-evident that professional liability insurance coverage is intended primarily to protect against liability to individual patients and that there is a lesser risk of such liability associated with research activities than with a department's funtions which are more apt to involve patient care. Indeed, Vanderbilt did not seriously dispute this point. Rather, Vanderbilt argued that the insurance costs were properly allocated through the DA cost pool as costs benefitting "both sponsored research and the patient care and other activities of the Medical School." Vanderbilt's Opening Brief, p. 88.
While we find, as explained below in part C., that these insurance costs benefit organized research in general, not just those projects involving patient care or otherwise involving human subjects, we also find that the use of the MTDC base is inappropriate given the lesser risk associated with research activities. We cannot say that these costs are of "like nature [with other DA costs] in terms of their relative contribution . . . to the particular cost objectives. . . ." OMB Circular A-21, Section E.2.b. We find, having accepted the Agency's basic premise, that the allocation of insurance premiums on an MTDC basis would thus produce an overallocation to organized research.
Furthermore, the Agency persuasively explained that, in the allocation of costs in the DA cost pool, each MTDC dollar associated with the organized research function is in effect being burdened by the same amount of insurance costs as each MTDC dollar associated with the University's other functions, notwithstanding organized research's lesser involvement with patient care-related activity. Yet, organized research should logically bear some lesser burden of insurance costs to reflect the obviously smaller risk of malpractice claims arising from research activities. Besides a general failure to recognize this lesser risk, there is, moreover, an averaging effect from the use of the MTDC base to allocate the disparate premium amounts that would shift costs to organized research. Tr., Vol. 3, pp. 524-525. Vanderbilt in our view never meaningfully addressed this overallocation, instead concentrating in its arguments on its positions that the rating and modifier system itself was fair and that the small amount of insurance costs to be allocated did not warrant the formation of a separate cost grouping, as advocated by the Agency. E.g., Tr., Vol. 3, pp. 499-504.
C. The Agency's Suggested Approach to Allocating Liability Insurance Premiums
While the Agency argued that the use of an MTDC base produces an overallocation to organized research, the Agency accepted Vanderbilt's general point that liability insurance costs provide some benefit to organized research. Agency's Brief, p. 48. Nevertheless, the Regional Director disallowed all liability insurance costs allocated to organized research, since in his view Vanderbilt had not proposed an acceptable methodology for their allocation. Regional Director's letter, p. 9.
The Agency in its brief refined this position and itself proposed a formula for allocating insurance costs to organized research. Agency's Brief, p. 50. The Agency proposed that liability insurance costs be treated as a separate cost grouping, rather than being included in the DA cost pool. The Agency proposed applying the following formula to this separate grouping of insurance costs:
Total Direct Salaries and Wages Charged to Organized Research Projects Involving Human Subjects Total Direct Salaries and Wages (Including Incentive Pay from VPPP)
Id.
The Agency's rationale for examining only research projects involving human subjects was that, according to the Agency, these are the only research projects that give rise to the risk of malpractice claims. Id. The Agency looked at salaries and wages because it considered this to be a good indication of the time spent on a research project. Id.
Vanderbilt rejected the Agency's approach here for two reasons. Vanderbilt argued that first, liability insurance costs are not material enough to warrant a separate cost grouping and, second, in any event, the formula proposed by the Agency was based on false assumptions, most importantly that the risk of liability from organized research arises only from those projects involving human subjects. Vanderbilt's Reply Brief, pp. 48-50.
We do not find merit in the University's first objection that the allocation of liability insurance costs may not warrant a separate cost grouping. We found above that there is an overallocation to organized research when these costs are included in the DA cost pool. Vanderbilt argued that these costs should not be treated as a separate grouping because OMB Circular A-21 describes, as .one factor in determining the appropriateness of using a separate cost grouping, that:
The number of separate cost groupings within a category should be held within practical limits, after taking into consideration the materiality of the amounts involved and the degree of precision attainable through less selective methods of distribution.
Section E.2.c.(6); Vanderbilt's Opening Brief, pp. 102-103. Vanderbilt maintained that the liability insurance costs here were not material. Vanderbilt's Opening Brief, p. 103. Insurance costs for FY 1983 represented .23 percent of total indirect costs allocated to organized research and, for FY 1984, .98 percent. Id. The Agency argued that these costs were material, noting that the total amount allocated for FY 1984 was $159,000, a significant figure in absolute terms, comprising approximately one half of one percent of the total indirect cost rate for a major institution. Tr., Vol. 3, p. 529. (The disallowance for FY 1984 was later recalculated to be $139,000.)
Although the parties disagreed about whether the costs here were "material," neither party offered an objective standard for determining whether costs are material in a particular case. We find, however, that we do not need to decide upon or apply such a standard, since we do not believe that OMB Circular A-21 intended the materiality of costs to be the sole criterion used in deciding on the need for a separate cost grouping. The introductory paragraph in section E.2.c. of the Circular provides:
General considerations on cost groupings.
The extent to which separate cost groupings and selective distribution would be appropriate at an institution is a matter of judgment to be determined on a case-by-case basis. Typical situations which may warrant the establishment of two or more separate cost groups . . . within an indirect cost category include but are not limited to the following . . . [including section (6), quoted above].
This language provides that the need for a separate cost grouping is a discretionary "judgment," to be determined "case-by-case" depending on a range of possible factors. Furthermore, paragraph E.2.c.(6) itself provides that the "materiality" of costs should be considered in conjunction with the "degree of precision attainable through less selective methods of distribution."
We find that the distortion that results from including liability insurance costs in the DA cost pool and applying an MTDC base results in a sufficiently imprecise method of allocating these costs to warrant their separate treatment. Moreover, we conclude from the record here that these costs are readily identifiable; Vanderbilt did not explain why the separate treatment of these costs would itself present some administrative burden that would make unreasonable the use of a separate cost grouping.
While we agree that a separate cost grouping may be an acceptable way in which to allocate these costs, we do not think it warranted, on the basis of the record presented, to preclude entirely the possibility of including liability insurance costs in the DA cost pool, as long as some type of adjustment is made for the distorting effect of application of the MTDC base. As to Vanderbilt's second objection to the Agency's approach, we agree that the Agency's formula is inequitable because of its assumption that the risk of liability claims would arise only from research projects involving human subjects. In its brief and appeal file, Vanderbilt explained that there is an exposure to liability claims from projects other than those involving patient care. The University submitted a letter from its insurance broker which stated:
. . . The carriers writing medical malpractice insurance do consider a physician doing research only to have a medical malpractice exposure. Accordingly, the carriers do make a premium charge for such physicians. In regard to your program, the charge is reflected in the overall premium. . . .
Vanderbilt's Att. IV.A.
The Agency presented no substantial response to Vanderbilt's explanation that the risk of liability exists with research projects not directly involving human subjects. At the conference, the Agency itself appeared to retract its earlier, more extreme view that liability insurance premiums benefit only research projects involving human subjects. Tr., Vol. 3, pp. 508-509. We also note that, while the disparate premium amount for research physicians (i.e., the 50 percent discount) supports the conclusion that a lesser risk is associated with research, for this purpose the distinction is drawn based on the amount of research, with no further discount where no human subjects are involved.
Another objection by Vanderbilt to the Agency's formula was the inclusion in the denominator of total direct salaries and wages, including incentive pay from the professional practice program. Vanderbilt's Reply Brief, p. 51. The Agency explained here that it looked at salaries and wages because they are "a reflection of the time spent on a sponsored project and thus measure the degree of exposure to the risk." Agency's Brief, p. 50.
Vanderbilt argued that salaries and wages are not a reliable indication of the degree of exposure to the risk of liability since, for instance, one exposure to a particular research project is enough to create a risk of liability and the need for liability insurance. Vanderbilt's Reply Brief, p. 51. At the conference, the Agency did not respond to this point. Since we found the Agency's formula inequitable on other grounds, we do not reach the more general issue of whether a salary and wage base might be appropriate for the allocation of liability insurance costs. In any event, the parties may, of course, explore further whether to use a salary and wage base.
Conclusion
In sum, the parties should develop some mutually acceptable approach for allocating the costs of liability insurance to organized research, consistent with our findings above. While we have accepted the Agency's approach of treating the costs as a separate cost grouping, we would also view acceptable the inclusion of liability insurance costs in the DA cost pool, as long as some weighting factor were applied to adequately account for the lesser likelihood of liability associated with organized research. If the parties are unable to agree, they may return to the Board on this issue.
IV. The Vanderbilt Professional Practice Program_(VPPP) Business Office
The Vanderbilt Professional Practice Program consists of Medical School faculty who maintain a practice in the Vanderbilt University Hospital and affiliated hospitals. The VPPP Business Office performs the billing and revenue collection activities for the VPPP. In its proposals for FY 1983 and FY 1984, Vanderbilt included the costs of the VPPP Business Office in the Departmental Administration (DA) cost pool, for allocation to all major functions of the University, including organized research. The Agency maintained that the costs of the VPPP Business Office should instead be directly allocated to the VPPP itself, and it disallowed $783,000 for FY 1983 and $841,000 for FY 1984.
As explained below, we uphold the Agency's disallowance. After explaining our conclusion about the allocability of VPPP Business Office costs, we then consider an alternative argument made by Vanderbilt that, if these costs are not allocable to the other major functions of the institution, the Board should require that the salaries and wages received by the faculty for participation in the VPPP should be excluded from the modified total direct cost (MTDC) base used in allocating the DA cost pool for the Medical School. As explained further below, we reject this alternative argument.
A. The Issue of Inclusion of VPPP Business Office Costs in the DA Cost Pool
The University's stated rationale for including VPPP Business Office costs in the DA cost pool was that the income generated from the VPPP program was used for more general purposes of the University than to merely provide supplemental income to faculty because of their hospital practice, so that VPPP Business Office costs benefitted and should be allocated to all functions of the University. Vanderbilt's Opening Brief, p. 108. Vanderbilt argued that in addition to generating income, the professional practice program created an undefined but nevertheless real benefit to the University as a whole, including the instruction and research functions, by, in effect, ensuring the existence of a clinical experience for students and researchers. Tr., Vol. 3, p. 541.
The Agency noted that Vanderbilt's argument concerning the benefit provided by the VPPP Business Office was "extremely strained" and argued that, if the only benefit derived from the VPPP Business Office was the raising of revenue, this would constitute an unallowable "fund raising" activity under OMB Circular A-21. Agency's Brief, p. 52. The Circular provides in section J.17.b.:
Costs of organized fund raising, including financial campaigns, endowment drives, solicitation of gifts and bequests, and similar expenses incurred solely to raise capital or obtain contributions, are unallowable.
An apparent rationale for the Circular's prohibition on the allowability of such costs is that fund raising is primarily intended to enhance general University revenue and is inherently not sufficiently related to the purposes of a grant or contract. Vanderbilt here did not attempt to refute this argument of the Agency. Vanderbilt's Reply Brief, pp. 53-60. We do not, however, decide whether the Business Office costs should be excluded from the DA cost pool on the basis of the prohibition of fund raising costs since we sustain the Agency's objection to inclusion of these costs on the grounds discussed below.
As to Vanderbilt's more fundamental argument about an incidental benefit to the instruction and organized research functions, the Agency argued that Vanderbilt's inclusion of these costs in the DA cost pool violated OMB Circular A-21's basic rules regarding allocability. Tr., Vol. 3, pp. 547-549.
OMB Circular A-21 provides in section C.4.a. that:
A cost is allocable to a particular cost objective (i.e., a specific function, project, sponsored agreement, or the like) if the goods or services involved are chargeable or assignable to such cost objective in accordance with relative benefits received or other equitable relationship.
* * * *
The Circular also provides:
Where a cost grouping can be identified directly with the cost objective benefited, it should be assigned to that cost objective.
Section E.2.d(2).
The Agency argued persuasively here that the relevant "cost objective" under Circular A-21 for the costs of the VPPP Business Office is the VPPP itself, rather than departmental administration. Tr., Vol. 3, p. 549. Vanderbilt did not deny the fact that the VPPP Business Office provided no administrative services to activities other than the VPPP. Vanderbilt's Reply Brief, p. 54. We therefore conclude that the costs of the VPPP Business Office should have been assigned to the VPPP itself.
We further find that Vanderbilt's proposal to include those costs in the DA cost pool is inequitable and would undoubtedly create a substantial overallocation to organized research. OMB Circular A-21 defines "departmental administration expenses" as follows:
The expenses under this heading are those that have been incurred for administrative and supporting services that benefit common or joint departmental activities or objectives in academic dean's offices, academic departments and divisions, and organized .research institutes, study centers, and research centers. * *
* * Section F.4.a. While Vanderbilt maintained that the VPPP provides an incidental benefit to the other functions of the Medical School, Vanderbilt did not in our view demonstrate that the VPPP Business Office costs were themselves of such a nature as to "benefit common or joint activities or objectives" of the Medical School, nor were the costs incurred to specifically benefit the work of "academic dean's offices, academic departments, or organized research institutes, study centers, and research centers."
To reiterate our point above, the VPPP Business Office performed administrative activities only for the VPPP, and not the other operations of the Medical School. Even if we accepted Vanderbilt's general point that the VPPP provided an allocable benefit to the other major functions, the solution would not be to include these costs in the DA cost pool. To do so would overallocate costs to organized research, since it would allocate the Business Office costs on an MTDC basis to each major function, which would distribute the costs proportionately to each of the Medical School's functions. Unlike the other types of DA costs, however, such as the operation of the dean's offices, the VPPP Business Office does not provide equal support to all the major functions, under Vanderbilt's own description of the VPPP and the Medical School.
The Agency cited as analogous support OMB Circular A-21's treatment of the costs of Sponsored Projects Administration (SPA). Tr., Vol. 3, p. 559. The Circular provides that the costs of SPA, in the absence of some alternative methodology demonstrated by the institution, be allocated "to the major functions of the institution under which the sponsored projects are conducted on the basis of the modified total cost of sponsored projects." OMB Circular A-21, section F.5.b. Since the vast majority of sponsored projects are conducted for the organized research function, a large part of these administrative costs would be allocated to the organized research function.
The example of the SPA cost pool admittedly does not provide a flawless analogy for the Agency's position, since the SPA cost pool is not simply directly charged to sponsored projects, but is allocated on an MTDC basis to those functions under which the sponsored projects are conducted. The Agency here argued only that the costs of the VPPP should be directly charged to the VPPP. We find that the Circular's treatment of the SPA cost pool does, however, illustrate the Agency's point that the costs of the VPPP Business Office should not be included in the DA cost pool solely on the basis that they are administrative in nature. The Circular clearly provides that costs may be allocated only to benefitting functions. As with the costs of the VPPP Business Office, SPA costs cannot be included in the DA cost pool, allocated to all major functions of the institution, when the only function deriving a benefit is sponsored projects.
Vanderbilt in our view had no substantial response to the Agency's analogy to the SPA cost pool. One distinction which Vanderbilt offered here was that the VPPP was a "discrete" activity which was separate from the Medical School in a way in which Sponsored Projects Administration was not. Tr., Vol. 3, pp. 582-584, 589. Vanderbilt also appeared to maintain that the "spillover" benefit from the patient care function was in practice greater than with sponsored projects. Tr., Vol. 3, pp. 590-591.
We fail to see the relevance of the degree of separation between the VPPP or the SPA activity from the rest of the Medical School to our analysis of the benefit received from these costs. If anything, the degree of separation between the VPPP and the rest of the Medical School argues for assigning the Business Office costs directly to the VPPP. As to Vanderbilt's apparent argument that VPPP produced some greater incidental benefit than SPA costs, Vanderbilt presented no evidence that this was so. We would expect that the existence of research activities at an institution would also provide some incidental benefit to the institution as a whole, for instance in terms of the reputation of the University and the ability to attract highly qualified faculty and students. In any event, the degree of "incidental benefit" derived from the VPPP or from research activities does not relate to the ultimate issue here of whether the VPPP Business Office costs should be included in the DA cost pool. B. Vanderbilt's Alternative Argument About the MTDC Base
The University argued alternatively that, if the Board agreed with the Agency that VPPP Business Office costs should not be included in the DA cost pool, the Board should also find that $4,900,000 of faculty salaries and wages from their participation in the VPPP should be excluded from the MTDC base used to allocate the DA cost pool for departments of the Medical School. Vanderbilt's Opening Brief, p. 109. The effect of excluding these salaries and wages from the MTDC base would be to increase the total amount allocated to organized research, reducing the disallowance relating to the VPPP. For FY 1984, Vanderbilt calculated that this would reduce the disallowance relating to the VPPP from $839,000 to $339,025. Vanderbilt's Opening Brief, p. 118; Vanderbilt's Att. V.E. (Vanderbilt did not indicate what the precise amount of this reduction would be using the Agency's last statement of the amount in dispute for FY 1984).
The University's rationale for excluding VPPP salaries and wages from the MTDC base was that, if the Board agrees that the VPPP Business Office does not provide a benefit to the other major functions so as to warrant the inclusion of these costs in the DA cost pool (as the University in its primary argument maintained that it does), the Board must also conclude that the VPPP does not receive a benefit from the administrative services of the Medical School so as to warrant inclusion of VPPP salaries and wages in the base used to allocate DA costs. Otherwise, maintained the University, the disallowance would be based upon an inconsistent treatment of the costs of the VPPP. Vanderbilt's Opening Brief, p. 118.
While the University's argument here may have some intuitive appeal, the record before us does not bear out the University's concern. First, the primary issue, as we found above, was whether the VPPP Business Office provided a measurable benefit to the institution's other functions so as to warrant inclusion of its costs in the DA cost pool, under OMB Circular A-21. The issue of whether the VPPP Business Office received administrative support from the Medical School appears to us a separate factual question, with no necessary connection to whether the costs of the VPPP Business Office are allocable to other major functions. We therefore do not agree that it would be "inconsistent" to include VPPP salaries and wages in the MTDC base used to allocate DA costs.
Second, on the basis of the record, Vanderbilt has not demonstrated to us that VPPP activities do not indeed receive a benefit from DA costs. While the University stated conclusorily that the VPPP "derives only insignificant benefit from the DA cost pool," Tr., Vol. 3, p. 544, the Agency explained that the professional practice program was an "integral part" of the Medical School, along with research and instruction. Tr., Vol. 3, p. 562. We would logically expect that the same faculty who engage in research and instruction also often maintain a practice in the hospital. The University itself appeared to agree that the VPPP was integrally related to the Medical School through its position that the VPPP was an essential function of the Medical School and provided a wider benefit to the University than merely to service patients in the hospital.
The Agency, as an example of the administrative support received by faculty for their participation in the VPPP, noted that the job descriptions of the Dean of the Medical School and department heads indicated that they provided administrative oversight to faculty involved with patient care/hospital activities. Vanderbilt's Att. I.O., Ex. R. Also, the salaries which faculty receive for their participation in the VPPP are set by the appropriate department chairperson and approved by the Dean of the School of Medicine. Vanderbilt's Opening Brief, p. 120. The University disputed the significance of this oversight, alleging that the Dean of the School of Medicine and departmental chairpersons "do not provide measurable support to the physicians' direct patient care activity similar to that provided to instruction, research and other institutional activities." Vanderbilt's Att. I.O., p. 30; Vanderbilt's Reply Brief, p. 57.
The Agency also argued that the VPPP was similar to Sponsored Projects in terms of the administrative support received from the rest of the Medical School. Tr., Vol. 3, pp. 566-567. The Agency observed that, while sponsored research is a "unique service" that has "no relationship to the normal instructional operations or the patient care operation," the salaries and wages of faculty involved with sponsored research are certainly included in the MTDC base for distribution of the DA cost pool. Tr., Vol. 3, p. 567.
We find that the burden here rests clearly on Vanderbilt to demonstrate that the VPPP received no benefit from the DA cost pool and thus that VPPP salaries and wages should be removed from the MTDC base. This issue was not raised directly by the Agency, but was based instead on the University's analysis that the Agency was being "inconsistent" in its treatment of the VPPP. Furthermore, only the University would have full access to the information that might provide evidence to document this argument.
In support of its position, Vanderbilt argued that there is a separate administrative cost structure supporting the VPPP, Vanderbilt's Opening Brief, pp. 122-124; that VPPP "incentive earnings" are excluded from the University effort reporting system, id., p. 121; and that the VPPP Business Office is located in physically separate quarters from the rest of the Medical School, Tr., Vol. 3, p. 575.
As to the alleged existence of a separate administrative cost structure supporting the VPPP, Vanderbilt in its final argument at the conference stated that "[t]he VPPP business office operates totally on support . . . [e]xpenses paid from VPPP revenues. These expenses are separately budgeted and easily identifiable. No departmental administration expenses benefit the VPPP business office." Tr., Vol. 7, p. 848. While Vanderbilt may be correct that the VPPP Business Office receives no administrative support from the rest of the Medical School, this does not answer the question of whether the operation of the VPPP itself receives any such benefit.
Furthermore, we find that the evidence presented by Vanderbilt does not prove the existence of completely separate administrative support for the VPPP. The only one of the exhibits cited by Vanderbilt which appeared to address this issue was Vanderbilt's Attachment V.L., which Vanderbilt maintained indicated that "VPPP accounts are being used to fund patient care administrative support and current operating expenses." Vanderbilt's Opening Brief, p. 124. Vanderbilt in its brief alleged that the exhibit "details the administrative support type expenses funded from VPPP sources in support of patient care." Id. Yet, Vanderbilt's Attachment V.L., entitled "Schedule of Costs by Object Code, FYE June 30, 1984," is merely a two-page listing of "Object Code Numbers" and accompanying descriptions of various miscellaneous administrative-type expenses, with the total amount of costs associated with each category. Assuming that this exhibit demonstrates some significant amount of administrative activities which were funded by the VPPP, it does not address whether the VPPP also received significant support from the rest of the Medical School to justify inclusion of VPPP salaries and wages in the base used to allocate the DA cost pool.
As to Vanderbilt's argument about the exclusion of VPPP incentive earnings from the effort reporting system, we fail to see the relevance of this observation. The purpose of this exclusion would appear to derive from the simple fact that "incentive" payments, designed to encourage participation in the VPPP, are not associated with the actual "effort" by faculty which is usually reflected in salaries and wages. Vanderbilt in its brief explained that incentive earnings are separate from a faculty member's base pay, Vanderbilt's Opening Brief, p. 121, which apparently includes the patient care related salaries and wages which Vanderbilt argued should be excluded from the MTDC base. Vanderbilt's Opening Brief, p. 109.
We likewise do not find availing Vanderbilt's point that the VPPP Business Office is in a separate building from the rest of the Medical School. We note that the Business Office is also separate from the hospital itself. Vanderbilt's Conference Ex. 1 (Campus Map); Tr., Vol. 3, pp. 575-576. The location of the VPPP Business Office itself is clearly not relevant here. Given Vanderbilt's argument, as part of its primary position, about the interrelationship of the VPPP and the other functions of the Medical School, it appears logical to expect that the departmental administrative support for the Medical School would also benefit the day-to-day functioning of the VPPP.
In support of its secondary position that faculty salaries related to the VPPP should be excluded from the MTDC base, the University cited the Board's decision in another cost allocation dispute, Wayne State University, Decision No. 806, November 13, 1986. In that decision, the Board agreed with Wayne State University that salaries received by University deans and department heads for their work at local hospitals should have been excluded from the MTDC base in allocating DA costs. The University explained in that case that the work performed by deans and department heads was not University-related and indeed that the personnel were in effect paid by the hospitals themselves.
We find the Wayne State case distinguishable from the situation presented here by Vanderbilt. The "Salary Reimbursement from Local Hospitals" section of the decision in Wayne State was clearly confined to the particular facts of that case. Wayne State University presented evidence that the functions performed by the deans and department heads were indeed independent of the University and that they did not benefit from DA costs at the University. In Wayne State the Board found it dispositive that the Agency had not rebutted the University's specific assertion that no DA services were required for the duties of the deans and department heads at the local hospitals.
In the case at hand, by contrast, the Agency has explained how the patient care activities provided by faculty at the Vanderbilt University Hospital were a part of the overall function of the Medical School, along with instruction and organized research. Indeed, the University acknowledged that the patient care function was a critical part of the faculty's duties at the Medical School. As we concluded above, the record in the present appeal causes us to conclude that Vanderbilt has not demonstrated that faculty received no support from general administrative operation of the Medical School for their patient care activities as part of the VPPP. For these reasons, we find the case at hand to be distinguishable from the Wayne State appeal.
C. Conclusion
In conclusion, we uphold the Agency's disallowance, since we agree that, under the standards of OMB Circular A-21, the costs of the VPPP Business Office did not provide a benefit to all functions of the institution so as to be included in the DA cost pool. We also reject Vanderbilt's alternative argument that $4,900,000 of faculty salaries and wages from their participation in the VPPP should be excluded from the DA cost pool, since the record does not support Vanderbilt's argument that the functioning of the VPPP was not supported administratively at the departmental level in the Medical School.
V. Student Services Costs
For the years at issue, Vanderbilt allocated a certain percentage of all "student services" costs to organized research. Student services include activities such as student health services, student counseling, campus recreation, and the admissions, registrar, and financial aid offices. OMB Circular A-21, section F.7.; Tr., Vol. 3, p. 608. For each year Vanderbilt allocated student services costs using a formula which compared the total work effort by students on organized research to the total student effort on campus. The Agency disallowed all costs allocated to organized research in this manner. The Agency disallowed $197,000 for FY 1983 and $139,000 for FY 1984.
The basis for the Agency's disallowance was two-fold. First, the Agency objected to Vanderbilt's allocating some part of all student services costs to organized research, since not all student services costs are sufficiently related to a student's functioning in his or her role as an employee at the University. Second, the Agency objected to the particular formulas chosen by Vanderbilt, essentially because they relied upon what the Agency asserted were unreasonable assumptions regarding student activities at the University. See Regional Director's letter, p. 12.
We find that the Agency's objection that not all student services costs should comprise the pool of costs to be allocated to organized research is well-founded. As to the particular formula employed, we find that some of the Agency's objections have merit but find for Vanderbilt as to some of the other elements. The parties should determine the total amount of student services costs allocable to organized research as a result of our analysis below.
A. The Cost Pool to be Distributed
OMB Circular A-21 provides a standard methodology for the allocation of the costs of student administration and services. Section F.7.b. provides that such costs be allocated to the instruction function, in the absence of a cost analysis study, unless "it can be demonstrated that the use of a different base would result in a more equitable allocation of the costs. . . ." Section E.2.d.(4)(a). Vanderbilt chose to use a "different base" in allocating student services costs and relied upon the formula we discuss below.
The cost pool which Vanderbilt distributed using its formula was comprised of all student services costs. The Agency, as a matter of long-standing policy, has objected to this approach, since it believes that not all student services may be viewed as allocable to organized research, but only those which are similar to the fringe benefits or services received by an employee.
The Agency's view that not all student services costs are properly considered in allocating costs to organized research was expressed formally in a 1979 written interpretation of OMB Circular A-21. In a publication entitled "HEW Interpretations of OMB Circular A-21," the Department's Office of Grant and Contract Financial Management provided a series of "questions and answers" regarding various items of cost.
The publication provided on pages 18-19:
F.7. Student Administration and Services
The standard allocation base for student administration and services requires that these expenses be allocated entirely to instruction. If an institution wishes to use a cost study or alternate base, under what conditions can these expenses be allocated to organized research?
An allocation of these expenses to organized research would be accepted where an institution can show that a given service benefits organized research. These services would be those associated with students performing substantive work on organized research projects which are analogous to fringe benefits or services normally associated with employees (e.g., health services, personnel administration, etc.). The burden of proof for establishing that substantive work was performed by the students rests with the institution. However, such services would not be allocable to organized research if they duplicate fringe benefits or services provided to student-employees in their capacity as employees.
Agency's Ex. B.
The language here that was the subject of dispute was that student services should be "analogous to fringe benefits or services normally associated with employees." The University argued that this concept imposed an additional requirement to that provided by OMB Circular A-21 and that as such it was not binding on the University. One of Vanderbilt's arguments was that the Department of Health and Human Services lacked the authority to provide additional restrictions to an issuance by OMB that was designed to be applied government-wide.
We reject Vanderbilt's argument here that all student services may comprise the cost pool to be allocated, because we do not agree that the Agency's written interpretation imposed an additional requirement. We therefore do not address the parties' arguments of whether the University would be bound to comply assuming the interpretation imposed some legal restriction not contained in the Circular.
Even if no such "interpretation" existed, we would agree with the Agency that not all student services provide a benefit to organized research through student employment of a type which justifies allocation to organized research. The Circular provides for a standard allocation of all such costs to the instruction function; the interpretation does no more than state the obvious point that costs distributed to organized research must be properly allocable to that function, i.e., they must provide the type of benefit contemplated by the Circular. If student employment on organized research projects is the justification for including organized research in the distribution base, then only costs analogous to fringe benefits or services normally associated with employees can be included in the pool of costs to be distributed.
While research may be an important function at some universities, we agree with the Agency that the primary purpose for which students attend a university is in order to obtain an education. We understand that certain graduate students may intermingle their work on a dissertation with a paid research position at the University. However, the implication of the University's approach to allocating student services costs is that students somehow serve on an equal basis in each of their roles and thus that all student services should be viewed as equally supportive of both the student and employee functions.
One problem we see with the University's view here is that these student services were clearly created to provide a benefit to students in their traditional role in receiving an education. Indeed, a significant part of the student population at Vanderbilt are undergraduates, whose research role we would expect to be relatively small. It is our understanding that, assuming that no students worked on organized research projects at the University, almost all the same student services would be provided on campus, and the record does not contain any information to the contrary. Any incidental "benefit" received by organized research projects from student services is thus largely fortuitous.
On the other hand, we recognize that an apparent rationale for allocating some student services costs to organized research projects is that, if students were not available to perform this work, the researchers would need to hire additional employees for whom they would need to provide certain services beyond their salaries and wages. Under the Agency's interpretation, it is appropriate to use research funds to pay some share of student services costs even though the students receive these services because of their attendance on campus. In other words, because there is an advantage to research projects because of the availability of these student workers, one should regard that part of student services equivalent to fringe benefits as providing some benefit to organized research projects.
We therefore find reasonable the Agency's view that those student services which should form the pool of costs allocable to organized research are those analogous to fringe benefits and other services provided employees.
B. The Formula to Apply to the Pool of Student Services Costs
Vanderbilt in its briefs described two different formulas for FY 1983 and for FY 1984, which resulted in the allocation of very different amounts of student services costs to organized research. The Agency objected to the use of both these formulas, and also argued that Vanderbilt in the course of negotiations over the FY 1983 indirect cost rate had conceded that student services costs were not allocable for that year. See Agency's Brief, pp. 61-66, 71-72. At the conference before the Board, the parties agreed that Vanderbilt's final proposal concerning FY 1983 was to use the formula described as the FY 1984 proposal, and thus that the significance of the "FY 1983 formula" was moot. Tr., Vol. 5, pp. 460-463.
The one formula that was thus in dispute for both FY 1983 and FY 1984 was described by Vanderbilt as follows:
FTE of students working percent of student services on sponsored research = costs allocable to sponsored FTE students enrolled research
Vanderbilt explained how the numerator and denominator of this formula were calculated:
1. An FTE for students working on sponsored research was calculated as follows:
a. Total student salaries and wages allocable to sponsored research (derived from work-study effort reports) were divided by the average.hourly wage for students ($4.02) to calculate student hours worked.
b. FTE of students working was calculated by dividing student hours worked by 780 hours (39 weeks x 20 hours/week).
2. An FTE for students enrolled was calculated by dividing total student credit hours by 15 credit hours.
* * * *
Vanderbilt's Conference Ex. 28; see Vanderbilt's Conference Ex. 38 (applying methodology to FY 1983 data).
The Agency had a series of criticisms of this formula, which we present in turn below, followed by our findings on each point. In an exhibit prepared for this appeal, the Agency applied its own proposed formula to some rough data for FY 1983 to illustrate how it thought a proper formula would work. See Agency's Ex. PP.; Tr., Vol. 6, pp. 561-562.
The Agency first criticized Vanderbilt's method of computing total student salaries and wages allocable to organized research. Tr., Vol. 6, pp. 556-557. The University measured the allocation of work-study salaries and wages to the major functions of the University to serve as a sample for the distribution of campus-wide student salaries and wages. Vanderbilt maintained that the data did not exist to directly measure the allocation of all student salaries and wages. Tr., Vol. 6, p. 565.
The Agency first objected to Vanderbilt's approach here because the amount of work-study salaries and wages was relatively small compared to all student salaries and wages. Only $131,415 of salaries and wages were paid to students for their time working through the work-study program, compared to total student salaries and wages of $3,710,338 for FY 1983. Vanderbilt's Conference Ex. 38.
Further, the Agency found that Vanderbilt did not demonstrate how the distribution of work-study salaries and wages was a reliable surrogate for the distribution of total student salaries and wages. In this regard, Vanderbilt submitted that "[c]ollege work-study funds are used to support all direct cost functions within the University, including sponsored research, and therefore show a direct benefitting relationship." Vanderbilt's Opening Brief, p. 157. Vanderbilt also noted that "these were the best figures we had to approximate the effort allocation percentages for students who work." Tr., Vol. 6, p. 565.
While maintaining its argument about the effectiveness of work-study data as a base for distributing all student salaries and wages, the Agency also noted that the appeal record reveals in another context that Vanderbilt apparently was indeed able to measure the allocation of student salaries and wages to the major functions of the University, at least for all undergraduates. Tr., Vol. 6, pp. 583-585. Along with its arguments on the allocation of library costs, Vanderbilt submitted material which, in Vanderbilt's own explanation,
. . . identifies the Undergraduate student salaries and wages which were included in the salary and wage allocation base used in the FY 1984 indirect cost proposal.
Vanderbilt's Att. II.I. In response to a request by the Agency that it clarify the significance of this data, Vanderbilt at the oral conference explained that this exhibit did not reflect a complete analysis of undergraduate employment and that what was presented was the outcome of a very laborious effort. Tr., Vol. 6, pp. 585-586.
We do not believe that the record here is complete enough to allow us to fully evaluate the Agency's argument that Vanderbilt should have been able to perform a complete analysis of student salaries and wages. We do find, however, that Vanderbilt has not presented any evidence to the effect that the distribution of $131,000 in work-study salaries and wages would be representative of the total student salaries and wages of over $3,700,000. We therefore direct that, to the extent that Vanderbilt reasonably can do so, it should investigate the actual distribution of total student salaries and wages in arriving at a figure for the percentage of total student salaries and wages allocable to organized research. If it is unable to do so, then Vanderbilt should supply convincing evidence that the study of work-study salaries and wages provides a reliable indication of the distribution of all student salaries and wages.
The Agency's next objection to Vanderbilt's formula was the use of a 20-hour work week in the numerator. Tr., Vol. 6, pp. 557-558. We reject this criticism. The stated basis for the Agency's objection here was that a "full-time equivalent working is 40 hours . . ." and therefore "that's the acceptable number." Id. However, Vanderbilt's purpose in choosing a 20-hour work week was not to make some comparison to the usual full-time equivalent employee, but to calculate a per hour equivalent for working students, which could serve as a proper comparison to a similar measure of "students enrolled" in the denominator of the formula. Vanderbilt's Conference Ex. 28. The Agency here may have been confused by Vanderbilt's perhaps unfortunate use of the phrase "full time equivalent" in the numerator and denominator of its student services allocation formula.
As Vanderbilt's Exhibit 28 explains, the "FTE of students enrolled" in the denominator of the formula was calculated by dividing total student credit hours by 15 credit hours, to arrive at some "per hour" equivalent for students enrolled. The numerator of the formula must therefore be calculated in an analogous fashion. Since "total student hours worked" in the numerator was determined by an examination of actual salaries and wages paid students (even if, under Vanderbilt's approach, through the examination of some sample of salaries and wages paid), this figure must be divided by the actual amount of hours students would tend to work at the University.
As Vanderbilt reasonably explained, students do not work 40 hours on average and, indeed, are prohibited from working more than 20 hours. Tr., Vol. 6, p. 566. Vanderbilt's use of an average 20-hour work week would thus seem an appropriate, even conservative, way to establish an average figure for all students (since many students of course might work less than 20 hours on average).
The Regional Director in his determination also criticized the use of a 39-week year, rather than a more usual 52-week year. Regional Director's letter, p. 16. We reject this criticism for the same reason we outlined above with regard to the use of a 20-hour work week. Since students would not necessarily be on campus for the summer and other vacation periods, a 52-week period may be an inappropriate standard to use in establishing a per hour average of work effort by students to be used in the numerator of the formula.
The Agency next criticized Vanderbilt's consideration, in the numerator of the formula, of only University employment and that it did not also include off-campus employment. The basis for this objection was that students who work off-campus would avail themselves of many of the same student services as students who work on-campus. Tr., Vol. 6, p. 566.
We reject this criticism for two reasons. First, the Agency in effect wants to increase the denominator of the formula that determines the "FTE of students working" on Vanderbilt's Exhibit 28. Rather than 780 hours, the Agency argued that the appropriate figure should be much larger than that, in order to account for the phenomenon of students working outside the University. See item 1.b of Vanderbilt's Conference Ex. 28. Yet, the numerator of the relevant fraction there was based on student workers' relative contribution to organized research compared to the other major functions that might be expected of student workers on campus. The Agency presented no evidence that off-campus employment could ever involve some contribution to the organized research function (or to instruction, for that matter). In other words, the effect of the Agency's proposal to include off-campus employment would be to distribute a pool of costs (devotion of effort to all major functions of the University) over a base that would include a category of costs (off-campus employment) having nothing to do with the pool being distributed.
The second reason we find the Agency's position regarding off-campus employment to be unreasonable is that student workers in jobs off campus would, we expect, likely be entitled to fringe benefits and other services, similar to some "student services" on campus. Since we have accepted the Agency's position that the relevant pool of student services costs which might be allocable to organized research are those analogous to fringe benefits and other services provided to employees, the Agency's approach here to include off-campus employment would appear to constitute a double counting of the "fringe benefits" these off-campus student workers would receive. Tr., Vol. 6, pp. 574-577.
The Agency's next objection was that Vanderbilt did not include consideration of student's time spent outside the classroom in calculating the "FTE of students enrolled" in the denominator of the formula. Tr., Vol. 6, p. 559. The Agency's position was that the formula should consider that a student spends three hours of study time for each hour of classroom time. Id. We reject this criticism.
While the Agency may be correct that one may reasonably expect students, as part of their studies, to spend considerable time outside of the classroom, it is also possible that student workers might devote some considerable time to organized research activities outside of the time of their paid employment. Vanderbilt explained that this would be particularly so for those graduate students who combine research projects for which they are paid with work on their dissertation. Tr., Vol. 6, pp. 580-581.
Vanderbilt also noted that the Agency's proposal to include time spent outside of the classroom may be wrong, given the methodology proposed by OMB Circular A-21 in another context to determine a full time equivalent for students. In the standard allocation base for the allocation of library costs, the Circular describes how to establish an FTE for students, which does not anticipate an examination of additional hours to those spent in class. See section F.6.b, discussed above in section I.
The Agency's last criticism of Vanderbilt's formula was that the denominator should not only include an "FTE of students enrolled," but must also include an FTE of students working. Tr., Vol. 6, p. 560. We agree with the Agency here. The formula's clear purpose is to compare the effort by students relating to organized research employment to all student effort, in order to establish the significance of organized research employment on campus. The inclusion of only student enrollment is indeed only a part of the picture, and a figure for total student employment must also be added in.
C. Conclusion
In sum, we uphold the Agency with respect to its position that only those student services costs analogous to fringe benefits or other services provided to employees should comprise the cost pool for allocation of student services costs to organized research. The following formula for the allocation of the pool of this part of student services costs reflects our findings on each disputed issue.
1. Total effort by students working on organized research should be calculated as follows:
a. Total student salaries and wages allocable to organized research, from actual count of population (if reasonably obtainable) or, based on convincing evidence of their representative nature, from work-study effort reports. This total should be divided by agreed upon average hourly wage for students to calculate student hours worked.
b. This figure of total hours should be divided by 780 hours (39 weeks x 20 hours/week) to calculate the effort by students working on organized research.
2. Total effort of students enrolled should be calculated by dividing total student credit hours by 15 credit hours.
3. The total effort by students working on organized research should be divided by the sum of the per hour measure of students enrolled plus the per hour measure of all students working, including work on organized research projects.
In arriving at a measure of total student salaries and wages allocable to organized research in point 1.a above, the University should either supply data on the actual total distribution of student salaries and wages or, if unable to do so, should produce convincing evidence that the study of work-study salaries and wages provides a reliable indication of the distribution of total salaries and wages.
If the parties encounter further dispute in arriving at an amount of student services costs allocable to organized research, they may return to the Board on this issue.
VI. The Finality of an FY 1983 Rate Agreement
In its briefs before the Board, Vanderbilt argued that the disallowances for FY 1983 should be overturned because they were based on the Agency's alleged breach of a "final" negotiated rate agreement for that year. Vanderbilt's Opening Brief, p. 166; Reply Brief, pp. 71-75. The Agency disputed that this FY 1983 rate agreement reflected its final determination of the actual indirect costs for FY 1983. The Agency argued that it never intended to complete its review of the actual FY 1983 indirect costs during the negotiations preceding the FY 1983 agreement and that Vanderbilt was clearly informed of this. Agency's Brief, pp. 70-71. The Agency also made a counter argument that one part of the FY 1983 rate agreement, concerning the allocability of student services costs to organized research, was final and thus that Vanderbilt could not challenge the Agency's disallowance on this issue for FY 1983 before the Board. Agency's Brief, pp. 68, 71-72. Because of the complicated factual background of this part of the dispute, the Board conducted an evidentiary hearing on this matter.
Based on the testimony presented at the hearing and on an examination of the many documents presented by both sides, we find that the FY 1983 negotiated rate agreement did not preclude the Agency from further review and adjustment of the FY 1983 actual indirect costs. We also find that Vanderbilt was not bound by the concession it made during the negotiations agreeing to exclude the allocation of student services costs.
The letter and attached rate agreement which Vanderbilt claimed established a binding agreement for FY 1983, Vanderbilt's Attachment I.C., do not themselves even refer to making the FY 1983 costs final. Assuming that it might be conceivable that there nonetheless could have existed some mutual understanding concerning the finality of a rate for FY 1983, we find that Vanderbilt has not demonstrated this to be so. As a part of the allegedly binding agreement, the Agency told Vanderbilt that it was planning to conduct an on-site review of FY 1983 actual costs; we find that the Agency clearly intended this review to be the basis for possible adjustments to the FY 1983 actual indirect costs. Assuming for the sake of argument that Vanderbilt had reasonably thought that the purpose of this on-site visit was to adjust some period other than FY 1983, there was still no binding agreement for FY 1983 since the parties would have had no common understanding in this regard.
A. Background
The type of indirect cost rate used at Vanderbilt during the period relevant to this appeal was a "fixed rate with carry-forward provisions." For this type of rate, the difference between the fixed rate and the actual rate for a given fiscal year constitutes a "carry-forward" amount that would be included in calculating a future period's fixed rate. The parties negotiate to set the "fixed rate" based on the "carry-forward" amounts and on an estimate of costs for a future period. Once set, a fixed rate does not change. Since any over or under recovery of indirect costs from a particular fixed rate as compared to the actual rate is carried forward it affects only the fixed rate set for a subsequent fiscal year. Vanderbilt's Conference Ex. 29 (Public Health Service Grants Policy Statement, p. 62); OMB Circular A-21, sections G.3-G.5.
As explained at the hearing, the University would submit an indirect cost rate proposal after the close of a fiscal year which would serve as the basis for negotiations between the University and the regional office of DCA. The object of the negotiations would be both to determine the final actual indirect costs for that fiscal year and to set the next fiscal year's fixed rate. Tr., Vol. 4, p. 53. In June of 1984, from the negotiations concerning the FY 1983 indirect cost rate proposal, DCA set a fixed rate for FY 1985. Tr., Vol. 4, p. 43; Vanderbilt's Att. I.C.
Vanderbilt argued that the Agency was bound to a final determination on FY 1983 actual indirect costs by the June 26, 1984 letter from the Acting Director of the regional office of DCA and the attached rate agreement, transmitted to Vanderbilt for the signature of the appropriate University official. Vanderbilt's Opening Brief, p. 5; Vanderbilt's Att. I.C. The first paragraph of the letter read as follows:
The original and one copy of an indirect cost Rate Agreement are enclosed. This Agreement reflects an understanding reached between your organization and . . . a member of my staff as a result of negotiations conducted by telephone on June 26, 1984. An onsite review will be made of your actual costs for the period ended June 30, 1983. Any findings as a result of the review will be made against your 1984 actual costs. This onsite review will be conducted in the fall of 1984.
The attached rate agreement did not refer to FY 1983, but instead set a fixed rate for FY 1985 and noted the fixed rate set the prior year for FY 1984.
B. The Bases for the Parties' Positions
Vanderbilt maintained that this letter and the attached rate agreement finalized the rate and costs for FY 1983 because of the following:
1. The parties had concluded telephone negotiations which Vanderbilt said it understood to have resolved the issue of a rate for FY 1983. In order to agree upon a final rate, Vanderbilt had in the course of the negotiations conceded the exclusion of costs for student services.
2. Vanderbilt had understood in previous years that its receipt of similar transmittal letters for those years' indirect cost rate agreements indicated the finality of those years' rates.
3. The fact that the attached rate agreement did not address the FY 1983 rate, according to Vanderbilt, indicated that FY 1983 was a closed year and that the rate was final and not subject to future adjustment.
4. In the cover letter from Vanderbilt sending back to the Agency Vanderbilt's copy of the signed rate agreement, Vanderbilt noted that "the negotiation resulted in finalizing indirect cost rates for Fiscal Year (FY) 1983 and fixing indirect cost rates for FY 1985." Vanderbilt's Att. I.D. The Vanderbilt official who was Director of the Office of Indirect Cost Studies in 1984 at one point in his testimony appeared to dispute the Agency's allegation that Vanderbilt had received oral notification by telephone in September or October that the Agency did not consider the FY 1983 rate final. Tr., Vol. 4, p. 100. 5. While the June 1984 letter from the Agency referred to plans to make an on-site survey of FY 1983 "actual costs," the letter did not say that the outcome of such a survey might be to alter the FY 1983 rate or to take disallowances for that year. Instead, the letter indicated that "[a]ny findings as a result of the review will be made against [Vanderbilt's] 1984 actual costs."
6. According to Vanderbilt, after negotiations are concluded for a particular year, the only way in which the year can be "opened" is through an audit or because of fraud. Tr., Vol. 4, pp. 100-101, 168. Since the Agency's on-site visit did not constitute an audit and there was no allegation of fraud, the Agency had no basis on which to adjust the FY 1983 rate.
The Agency maintained that the FY 1983 negotiated rate agreement was based on a temporary determination as to FY 1983 actual costs, so that further adjustments were proper after the on-site review in the Fall of 1984, because of the following:
1. In addition to the reference to plans to conduct an on-site review in the June 26, 1984 letter, Vanderbilt had been informed by telephone before June 1984 that the Agency planned to make an on-site review of FY 1983 costs, which presumably could serve as a basis for adjusting the rates for that year. Tr., Vol. 5, p. 408.
2. In the previous year's negotiations, the Agency conducted an on-site review of actual costs for FY 1982 in the fall of 1983, the results of which were used to adjust the FY 1982 indirect cost rate and were reflected as a carry-forward amount in the FY 1985 fixed rate. See Vanderbilt's Conference Ex. 34; Tr., Vol. 4, pp. 86-87. While Vanderbilt protested this adjustment of the FY 1982 rate, Vanderbilt's Conference Exhibit 35, the University was clearly on notice of the Agency's purpose in conducting on-site reviews and thus should have understood the significance of the Agency's proposal to conduct an on-site review in the Fall of 1984.
3. According to the Agency, only the cover letter sent with the actual rate agreement might indicate whether a rate agreement was final for some past period. Tr., Vol. 5, pp. 352-355. Here, the June 26, 1984 cover letter nowhere indicated that the FY 1983 rate was final and indeed informed Vanderbilt that the Agency planned an on-site review of FY 1983 actual costs. While Vanderbilt maintained that cover letters for previous years that were similar to the June 26, 1984 letter indicated the finality of the past year's rates, this letter was unusual in its reference to an on-site review. Tr., Vol. 4, p. 116. In the immediately preceding year, the Agency had also expressed its intention to conduct such a review and indeed adjusted the FY 1982 rate on this basis.
4. The fact that the rate agreement attached to the June 1984 transmittal letter omitted mention of a FY 1983 rate did not indicate the finality of the FY 1983 rate agreement, as evidenced by rate agreements established with other universities in the same DCA region. The cover letters for agreements introduced into the record with two other Tennessee universities provided that the rates for the preceding fiscal year would not be final until after an on-site review was conducted. Yet, the attached rate agreements referred only to fixed rates for the two years after the one for which negotiations were conducted. As with the agreement sent to Vanderbilt in June 1984 (Vanderbilt's Attachment I.C.), the rate for the year whose actual costs were being negotiated (for Vanderbilt, FY 1983) had "dropped off" the agreement, but was clearly not yet final for those other two institutions. Agency's Exs. AAA, DDD; Tr., Vol. 5, pp. 320, 323-324.
5. Vanderbilt had itself placed the Agency in the difficult situation of establishing the fixed rate for FY 1985 in too short a period of time in which to conduct an on-site review, which the Agency, as a matter of policy, now required for "high-dollar" research institutions such as Vanderbilt. Vanderbilt's indirect cost rate proposal was originally due by the end of December 1983. Tr., Vol. 4, p. 127. On the basis of an extension granted by the Region, Vanderbilt submitted its original proposal in March 1983 and a revised proposal in late May 1984. Agency's Exs. O, P. In order to establish the fixed rate for FY 1985 by the start of the 1985 fiscal year on July 1, 1984, the negotiations had to be conducted entirely by telephone and an on-site review of FY 1983 costs was planned for the fall of the year, after a temporary rate was established.
6. The Agency's purpose in conducting an on-site review in the Fall of 1984 was clearly understood within the Agency. In a February 1983 memorandum to all regional directors of DCA, the Washington chief of DCA articulated a policy requiring on-site reviews for "high-dollar" research institutions. The memorandum provided:
Reviews of High Dollar Cost Proposals
Consistent with the workload standards, it is essential that thorough, usually on-site, reviews be made of indirect cost proposals from the "high dollar" universities . . . . If you conclude that a comprehensive review of one of these proposals . . . is not needed (e.g., no significant changes from the prior proposal and the prior proposal was subjected to a comprehensive review), the reasons must be documented in the institutional file.
Agency's Ex. EEE.
According to the regional director of DCA, the Agency had always planned to conduct an on-site review at Vanderbilt before finalizing FY 1983 rates, in compliance with the policy memorandum. Tr., Vol. 5, p. 287.
7. Until the time Vanderbilt filed its original brief before the Board, the University had never objected to the Agency's position that the FY 1985 fixed rate had been based on a temporary FY 1983 rate and that the Agency had the ability on the basis of the on-site review to adjust the FY 1983 rate and to ultimately effectuate disallowances for that year through carry-forward amounts for later fiscal years' fixed rates. Tr., Vol. 5, pp. 311-312. While the University had challenged the Agency's position regarding the finality of the FY 1982 agreement, Vanderbilt's Conference Exhibit 35, no such objection was repeated in any correspondence for the next year.
C. Analysis
The issue before the Board is whether the negotiations concerning the actual indirect costs for FY 1983 were "final" at the time DCA fixed the rate for FY 1985 on June 26, 1984. The Agency maintained that the only agreement the parties reached concerned the fixed rate for FY 1985; that rate needed to be set before the close of FY 1984 on June 30, 1984 to enable Vanderbilt to recover its indirect costs for FY 1985. According to the Agency, the second purpose of the negotiations, the establishment of the final "carry-forward" amount for FY 1983, had not yet been concluded. According to the Agency, Vanderbilt knew or should have known this, because the Agency clearly informed Vanderbilt that it planned to conduct in the fall of 1984 an on-site review of actual costs for FY 1983.
One basis for Vanderbilt's view that negotiations concerning FY 1983 were concluded was DCA's statement that any findings resulting from the on-site review would be adjustments to FY 1984 actual costs. Tr., Vol. 4, p. 49; Vanderbilt's Att. I.C. Vanderbilt maintained that it understood the Agency's reference to adjustments to 1984 actual costs to mean that the on-site review of Vanderbilt's FY 1983 actual costs would be an examination only of Vanderbilt's general methodology to provide a baseline for review of FY 1984 actual costs. Tr., Vol. 4, pp. 93, 154-155. As explained below, we find Vanderbilt's stated interpretation of the purpose of the on-site review to be unreasonable given what the record shows about how the rate-setting process operates. We thus conclude that if this was indeed Vanderbilt's understanding of the purpose of the on-site review, this understanding was inherently unreasonable and should not serve to bind the Agency to this interpretation.
The record shows that the purpose of an on-site review of the actual costs for a particular year would be to examine the allowability and allocability of that year's costs, with the possible consequence of adjusting the costs for that year. Tr., Vol. 5, pp. 291-292. Testimony from the Vanderbilt official was consistent since that official acknowledged that a review of actual costs would include allowability and allocability. Tr., Vol. 4., p. 117. The regional director of DCA testified that DCA would not "make an on-site review to just generally acquaint itself with the methodologies utilized by a particular university," Tr., Vol. 5, p. 291. The DCA regional director provided two reasons for this. First, actual costs and methodology must be considered "hand in hand," Tr., Vol. 5, p. 291-292, i.e., one cannot consider methodological concerns divorced in the abstract from actual costs. Second, a review to just examine methodology for the purpose of adjusting a future period's rate would be a waste of limited resources, since a university's proposed methodology could change once that year's indirect cost rate proposal is actually submitted, or the costs incurred for that later period might themselves change. Tr., Vol. 5, p. 292. Vanderbilt at the hearing did not appear to address these particular concerns.
The Agency established that, had it not intended the possibility of further adjustments to the FY 1983 actual costs, it would have just issued the standard rate agreement with no reference to an on-site review of FY 1983 costs. Tr., Vol. 5, p. 296. Also, since a fixed rate is not subject to later adjustment, the only way to effectuate further adjustments to the FY 83 actual costs would be as carry-forward amounts considered with FY 1984's actual costs to be reflected in the FY 1986 rate.
Vanderbilt appeared to argue that neither OMB Circular A-21 nor established custom provide for the use of a "temporary" actual rate in setting the fixed rate for an upcoming fiscal year with the possibility of a later review of the actual costs underlying the "temporary" rate. Tr. Vol. 4, pp. 177-178. We find, however that while the use of a "temporary" actual rate was not the Agency's usual practice, Vanderbilt did not show how this practice violated the Circular or any established, binding custom regarding the use of a fixed rate with carry-forward provisions. We further find on the basis of the record presented here, the use of such a "temporary" rate was unavoidable.
Vanderbilt did not contest the obvious ability of the Agency in its discretion to perform an in depth review of a fiscal year's costs. The record indicates that the Agency had no choice of when to conduct its on-site review, given the policy of verifying costs on site at "high-dollar" institutions. Moreover, the record shows that Vanderbilt was aware of the scrutiny then being given indirect cost recovery in general, and also aware of the practice of conducting such reviews prior to reaching a final agreement as to a fiscal year's costs. Tr. Vol. 4, pp. 117-120; see Agency's Ex. LLL (Report by the Comptroller General of the United States). Because of the timing of the receipt of Vanderbilt's indirect cost rate proposal in late May of 1984, the parties needed to quickly arrive at some type of agreement before the start of FY 1985 in order to set the fixed rate for FY 1985. Thus, Vanderbilt's own delay in submitting its FY 1983 indirect cost rate proposal, coupled with the need to both set a rate for FY 1985 and conduct an on-site review, necessitated the use of some type of interim determination of 1983's actual costs if there was to be a timely rate set for FY 1985. We find that the Agency adequately communicated its intent to further consider FY 1983 costs by informing Vanderbilt in its transmittal letter that an on-site review "will be conducted in the fall of 1984." The University could only reasonably have thought that negotiations were indeed ultimately not concluded for FY 1983, since one outcome of the on-site review might be to question areas of cost. Moreover, as noted by the Agency, Vanderbilt did not raise the issue of the alleged finality of an FY 1983 rate agreement until it was in litigation with the Agency before this Board.
We are not persuaded by Vanderbilt's point that the regional director of DCA did not clearly object to Vanderbilt's letter in which it characterized the status of an FY 1983 agreement as "final," before DCA had performed the on-site review. As explained above, DCA believed that it had already fully communicated its view of the temporary nature of its findings on the FY 1983 actual costs. When DCA received Vanderbilt's July 27, 1984 letter in which Vanderbilt described the FY 1983 rate as "final," the DCA regional director did indeed object to this description, according to his testimony. Tr., Vol. 5, pp. 303-304. Because of the press of business on the approach of a new school year, DCA did not have the opportunity to respond immediately to this comment in this letter received from the University. The DCA regional director testified that he spoke by telephone in early September with the Vanderbilt official in charge of indirect cost studies and explained that the FY 1983 rates were not final. We find his testimony to be credible and supported by other evidence in the record.
The Agency introduced into the record a copy of the University's July 27, 1984 letter, with the DCA regional director's handwritten notation, initialled on September 5, 1984, that he spoke by telephone with the Vanderbilt official in charge of indirect cost studies, explaining that FY 1983 rates were not final and might be adjusted because of the outcome of the on-site review. Agency's Ex. S. The note also mentioned that the Vanderbilt official offered that it was not necessary to record this conversation in writing. We have no reason to doubt the authenticity of this handwritten notation, which accords fully with the regional director's oral testimony. Testimony regarding this matter by the Vanderbilt official mentioned in the notation does not support Vanderbilt's position here. The official contradicted himself by first denying that DCA had ever responded to Vanderbilt's reference to "finalizing" the FY 1983 rate in the cover letter to the signed rate agreement, Tr. Vol. 4, p. 100, then by acknowledging, in response to fuller questioning from the Agency, that the Agency official had telephoned and asserted "that 1983 was still open." Tr., Vol. 4, p. 160. Although the Vanderbilt official recalled the telephone conversation as having occurred "closer to October," his testimony does not otherwise call the DCA regional director's testimony into question. The Vanderbilt official further testified to his understanding at the time of the on-site review that the Agency's object was to make further adjustments to FY 1983. Tr., Vol. 4, pp. 159-161.
We also find unavailing Vanderbilt's arguments that a negotiated agreement finalizing costs for a particular year may be reopened only by audit or because of fraud. The University cited no authority for this principle, either in terms of the indirect cost rate process or more generally in terms of principles of contract law.
Taking at face value Vanderbilt's stated understanding concerning the on-site review in the Fall of 1984 as a review solely of its methodology, we would conclude only that the parties lacked a common understanding as to the status of the FY 1983 costs. While it might ordinarily be inferred from a negotiated agreement that the prior year's costs had been finally established in order to set the new fixed rate, this is not expressly stated in the FY 1983 rate agreement and was not appropriately inferred given the provision for a later on-site review of FY 1983 actual costs stated in the transmittal letter. The correctness of our conclusion is underscored by Vanderbilt's failure to assert the finality of the FY 1983 rate until the Board proceeding. We thus conclude that the Agency was not precluded from further adjusting the FY 1983 actual indirect costs as a result of the on-site review conducted in the Fall of 1984.
D. Vanderbilt's Concession on Student Services Costs
The Agency made the counter argument that Vanderbilt was bound by a concession regarding student services costs it made in the course of negotiations over the FY 1983 rate. Agency's Brief, pp. 71-72. In order to reach some agreement regarding FY 1983 costs, Vanderbilt in telephone negotiations on June 26, 1984, apparently agreed to exclude these costs. Student services costs was an area of disagreement between the parties over a period of several years.
The Agency's position here regarding the finality of Vanderbilt's agreement on student services costs is to us directly at odds with the Agency view that the June 1984 agreement on FY 1983 actual costs was only temporary. The Agency in its brief, after making its point about the alleged finality of Vanderbilt's concession on student services costs, emphasized more generally that "the negotiation of Vanderbilt's FY 83 indirect cost rate proposal on June 26, 1984 was not intended to be a final and complete review of Vanderbilt's FY 83 indirect costs." Agency's Brief, p. 70. Since we have accepted above the Agency's argument regarding the overall lack of finality of the June 1984 agreement vis-a-vis FY 1983 actual costs, we see no reason to treat any concession regarding student services costs as any less temporary than the rest of the agreement. The regional director of DCA on cross-examination himself appeared to recognize this:
Q: So, is your testimony now that Vanderbilt's negotiation of its student services recovery was a temporary giving up of student services recovery?
A: That was a way of temporarily arriving at a fixed rate for '85.
Tr., Vol. 5, p. 341.
Even leaving aside the issue of the temporary nature of the overall agreement in June 1984 regarding FY 1983 actual costs, we find that Vanderbilt should not be bound on appeal by some position that was clearly taken in the course of negotiations to achieve an agreement on a fixed rate. Vanderbilt has never conceded its substantive position regarding the allocability of student services costs, which Vanderbilt explained before the Board it believed should be allocated based on the formula associated with its FY 1984 proposal.
E. Conclusion
In sum, while Vanderbilt may have been displeased about the proposed adjustments for the FY 1983 actual costs, the record does not support Vanderbilt's argument that the Agency was bound to a final FY 1983 rate by the negotiations in June 1984. The Agency clearly did not intend to agree that FY 1983 actual costs could no longer be considered, and communicated this to Vanderbilt. The written agreement and transmittal letter make no reference to the finality of FY 1983 costs and we do not accept Vanderbilt's view that the agreement should be construed to indicate that they were final. Finally, we reject the Agency's argument that one discrete part of the negotiations in question, that concerning the allocability of student services costs, should be viewed as final and binding on Vanderbilt.
VII. Vanderbilt's Concerns about Whether the Agency Acted in "Good Faith"
Vanderbilt argued that the Agency's disallowances for both years should be overturned because the Agency violated certain regulations governing procedures relating to indirect cost rate disputes. Essentially, Vanderbilt alleged that both DCA and the Regional Director failed to explain adequately the bases of each of their determinations and failed to negotiate in good faith with the University. Vanderbilt's Opening Brief, pp. 160-166; Reply Brief, pp. 76-78.
We first set forth the Department regulations cited by Vanderbilt. We then explain our conclusion that, under these regulations, Vanderbilt's arguments do not provide a legal basis for reversing the Agency's disallowances. We next explain our finding that, in any event, the documentary record amply demonstrates that the written determinations by DCA and the Regional Director afforded Vanderbilt a sufficient explanation of the reasons for the disallowances. We last discuss the issue of whether the Regional Director and DCA negotiated "in good faith" with the University, on which we make no particular finding.
Regulations Cited by Vanderbilt
Vanderbilt cited Department regulations governing "informal grant appeals procedures," which provide:
Where an agreement cannot be reached between the Director, DCA, and the grantee, the Director, DCA, will promptly notify the grantee in writing of the Director's determination. This notification will set forth the reasons for the determination in sufficient detail to enable the grantee to respond and will inform the grantee of its opportunity for reconsideration under this subpart.
45 CFR 75.4.
If the grantee receives an adverse determination from DCA and appeals to the Regional Director, "the grantee will be provided the opportunity to meet with the Regional Director to discuss the issue(s) and to submit additional information in support of its position." 45 CFR 75.6(b).
Vanderbilt also cited regulations which generally describe final decisions of the Department, which would then be appealable to this Board. Specifically, 45 CFR 74.304(a) provides:
Granting agencies and other Departmental components attempt to promptly issue final decisions in disputes and in other matters affecting the interests of grantees. However, they do not issue a final decision adverse to the grantee until it is clear that the matter cannot be resolved informally through further exchange of information and views.
Section 74.304(c) describes the contents of a final Agency determination:
(c) The decision is brief but contains-- (1) A complete statement of the background and basis of the component's decision, including reference to the pertinent statutes, regulations, or other governing documents; and (2) Enough information to enable the grantee and any reviewer to understand the issues and the position of the HHS component.
Section 74.304(d) describes certain particular language which a grantee could expect a final Agency decision to include, including how to appeal to the Board.
Finally, Vanderbilt cited the regulatory language which informs grantees that if correspondence from the Agency does not contain this material, it may not be a final determination:
If a decision does not contain the statement, information, and language described in paragraphs (c) and (d) of this section, the decision is not necessarily the granting agency's final decision in the matter. The grantee should notify the granting agency that it wishes a formal final decision following any further exchange of views or information that might help resolve the matter informally.
45 CFR 74.304(e). Analysis
Vanderbilt in our view has not demonstrated that any of these regulations would ever require this Board to reverse the Agency's determination. A reading of these regulations indicates that they are not intended to provide sanctions for Agency noncompliance and they certainly do not offer the remedy of reversing an Agency determination. They are clearly intended instead as general guidelines to inform grantees of how the Department's appeal processes operate.
The regulations in 45 CFR Part 75, specifically addressing indirect cost rate appeals, merely describe generally the process for appeals of indirect cost rate determinations. They inform grantees that both DCA and the Regional Director will seek to meet informally with a grantee before issuing a final determination, which will explain the reasons for any decisions reached, so that the grantee may respond.
The regulations at 45 CFR 74.304 inform grantees of what they can generally expect in the appeal of any preliminary Department determination. Sections 74.304(c), (d), and (e) merely inform grantees of whether correspondence they receive from a constituent agency indeed constitutes the agency's appealable final decision or whether the grantee should request a more formal determination from the agency after a further attempt at negotiation.
While the regulations contemplate that a final decision will provide the reasons for a determination, this Board has previously held that an agency may cure any inadequacies in a determination letter, so long as the grantee has sufficient opportunity to respond. See, e.g., New York Department of Social Services, Decision No. 274, March 31, 1982; Illinois Department of Public Aid, Decision No. 634, March 29, 1985. Furthermore, any alleged failure by the Agency to have fully and fairly explained the bases for the two preliminary decisions would have been cured by the extensive process provided to the University in its appeal before this Board. The Agency submitted a lengthy and comprehensive brief with numerous exhibits, which Vanderbilt responded to in its reply brief and over seven days of an oral conference. We have in this decision sought to respond to all of the University's arguments and other materials presented in support of its positions on the various complex issues raised by this appeal. While Vanderbilt may have found it troublesome that the Agency continued to refine some of its positions as this case developed, we consider such refinement inevitable given the nature and complexity of the issues here.
In any event, assuming that the regulations provided the type of remedy that Vanderbilt sought, we find that the written record does not support Vanderbilt's allegation that DCA and the Regional Director failed to adequately explain their decisions.
The Acting Director of the regional office of DCA issued a three-page letter which briefly described the basis for DCA's positions on each of the cost areas in dispute. Vanderbilt's Att. I.G. Vanderbilt did not explain how it was unable to respond to any of the items in the unilateral determination. The single example given by Vanderbilt as to the summary nature of the DCA regional director's letter was the director's explanation of the Agency's position regarding professional liability insurance, where the director stated that "it is the position of this office that professional malpractice liability insurance benefits only Patient Care." Vanderbilt's Att. I.G., p. 2, cited in Vanderbilt's Opening Brief, pp. 163-164.
However summary this explanation may seem, this was indeed the basic rationale for the Agency's disallowance. Furthermore, Vanderbilt in its brief failed to quote the next sentence of the unilateral determination, which offered that "until the University can develop a base that reflects the relative benefits derived to the related cost, these costs must be charged entirely to patient care." As we explained in our discussion of this issue in section III above, one basis for the Agency's position on this issue at this stage was that Vanderbilt had provided insufficient documentation to allow the Agency to verify the propriety of the University's approach. We therefore do not agree that the unilateral determination insufficiently explained the Agency's rationale.
Vanderbilt's complaint regarding the insufficiency of the unilateral determination is in our view further shown to be without merit by the fact that Vanderbilt did indeed prepare an extensive response to each of the points in dispute with DCA. In October 1985, the University prepared a detailed, 30-page report on its response to each of the cost areas. Vanderbilt's Att. I.K. While the report repeats Vanderbilt's view that the unilateral determination was issued without a full explanation and other documentary support, a review of this report indicates that Vanderbilt understood the basis for the Agency's position in each cost area and why the disallowance was taken. Vanderbilt did not allege that it found it necessary to obtain amplification of the basis for the determination prior to preparing its report. See Vanderbilt's Opening Brief, pp. 163-164.
We likewise reject Vanderbilt's assertion that the Regional Director did not adequately explain his decision on appeal from the unilateral determination. The Regional Director's June 27, 1986 determination was a 16-page single-spaced document that explored the basis for the parties' differing positions on each issue. Again, the University did not explain how the Regional Director's determination failed to disclose the reasons for the Agency's position on the issues. The University's 167-page opening brief before this Board would itself appear to belie this contention.
As to Vanderbilt's allegation that DCA and the Regional Director, in their negotiations with Vanderbilt, did not deal in "good faith," we make no particular finding. We denied Vanderbilt's request to present testimony, based on our preliminary conclusion that there was likely no legal basis for the remedy sought by Vanderbilt. We note, however, the inherent difficulty in assessing an allegation that the Agency (or a university) was unwilling to compromise or deal openly with another party at a particular series of negotiations. Moreover, as our analysis in sections I-V of this decision should indicate, there is no question that there existed in this case legitimate cost allocation disputes. A general expectation that parties negotiate in good faith certainly does not require that the parties always successfully conclude their negotiations, or that the parties avoid advancing their substantive disagreements for the sake of achieving a compromise. CONCLUSION
In conclusion, we uphold in full the Agency's disallowances for library costs, departmental administration non-labor costs, and costs of the Vanderbilt Professional Practice Program Business Office. Regarding costs of the Vanderbilt Professional Practice Program, we have also rejected an alternative argument by Vanderbilt that we should require the Agency to exclude faculty salaries and wages received for participation in the practice program from the MTDC base used to allocate the DA cost pool. We remand to the Agency the issues of allocating the costs of professional liability insurance and student services, in order to allow the parties to arrive at a reasonable allocation, consistent with our analysis in those parts of the decision. See our discussion on pp. 50-51 and 71-72 of those sections. In determining an allocation methodology for student services costs, we have directed Vanderbilt to investigate the distribution of total student salaries and wages, or, if it is unable to provide this information, to supply convincing evidence that the use of work-study salaries and wages is a valid substitute for all student salaries and wages. See pp. 67-68 of that section. Finally, we have rejected Vanderbilt's arguments about the binding effect of an FY 1983 rate agreement, as well as its other procedural arguments, and have also rejected the Agency's argument that Vanderbilt was bound to a concession made during negotiations to exclude the allocation of student services costs from its FY 1983 rate.
________________________________ Judith A. Ballard
________________________________ Norval D. (John) Settle
________________________________ Cecilia Sparks Ford Presiding Board