Despite increasing interest in interdisciplinary and translational research on the part of researchers and funding agencies,1–3 researchers have consistently cited institutional barriers as important deterrents.4–6 One major institutional barrier is the handling of the recovery of facilities and administrative (F&A) costs (i.e., “indirects” or “institutional overhead”),7–9 which have been traditionally allocated to the department or school of the principal investigator (PI).10 Researchers, administrators, and university ad hoc task forces and committees have suggested the development of processes for sharing F&A cost recovery as one way to minimize economic barriers to collaborative research across disciplines, departments, or academic institutions.5,7,9,11,12 In addition, in 2007, spurred by the Interdisciplinary Research Initiative of the National Institutes of Health (NIH) Roadmap, the NIH implemented a policy to allow multiple PIs on grants, as a mechanism to facilitate interdisciplinary and translational research and to maximize the potential of “team science” efforts.13
F&A costs are defined as “costs that are incurred for common or joint objectives and, therefore, cannot be identified readily and specifically with a particular sponsored project, an instructional activity, or any other institutional activity.”14 F&A cost recovery is typically awarded to institutions as part of funded research grants in excess of the direct costs required to conduct the research. For federally funded projects, F&A cost recovery rates, which vary among institutions, are governed by the Office of Management and Budget Circular A-21,14 are negotiated by a cognizant federal agency (e.g., the Department of Health and Human Services), and are reevaluated and updated periodically. Elements considered in the negotiation of an institution's F&A rate include the costs associated with multiple levels of administration as well as both the maintenance of the infrastructure (e.g., buildings, equipment, libraries, utilities) and the general operation of the physical facilities (e.g., heat, water). These F&A costs are then recovered by institutions at the negotiated rate, as funds awarded above the direct costs of conducting the project. For example, if a grant requesting $100,000 in direct costs is awarded to an institution that has a 61% F&A rate, the actual budget awarded would be $161,000: the $100,000 directly required to conduct the research and $61,000 to recover the associated indirect costs incurred by the project. But each individual institution determines whether a school, department, division, faculty member, or other entity actually shares in this cost recovery once awarded.
In 2005, the NIH issued a “Request for Information on the Plan to Recognize Multiple Principal Investigators,”15 which called for input regarding the allocation of funds. Although responses clearly indicated that budgetary issues would be an institutional barrier to the multiple PI leadership model,16 final NIH guidelines do not require the sharing of F&A cost recovery between PIs, schools, or departments but, rather, state that the decision to share (or not to share) F&A cost recovery should be determined “according to existing policy.”17 As a result, many institutions faced the administrative burden of developing such a policy for the first time18; however, despite the need for new policies, very little in the published literature focuses on the types of models in place or on institutions' experiences with them.
Thus, the purpose of this study was both to describe models for sharing F&A cost recovery for research conducted in universities and academic health centers (AHCs) and to gauge users' experiences with them. Using data obtained via policy statements available on the Internet for some institutions and data gathered directly from surveys of researchers and grants administrators from a range of institutions, we have described and categorized a number of different systems currently employed throughout the United States, as well as user perceptions and satisfaction as reported via our survey.
From November 2009 to January 2010, we conducted extensive searches using Google to locate policies on the distribution of F&A cost recovery posted publicly on institutional Web sites. We used 48 unique phrases, including the following:
* sharing/distribution of/allocation of indirect/facilities and administrative/F&A cost recovery,
* indirect cost/facilities and administrative/F&A distribution policy,
* sharing indirects/indirect cost recovery/facilities and administrative cost recovery/F&A/F&A cost recovery between departments/units/centers/schools,
* sharing of indirects/facilities and administrative cost recovery/F&A cost recovery to encourage interdisciplinary/multidisciplinary research,
* sharing indirect cost/facilities and administrative/F&A reimbursement,
* indirect cost/facilities and administrative/F&A distribution policy, and
* administrative barriers/barriers to interdisciplinary/multidisciplinary research.
In addition, using the same phrases, we performed searches directly on the Web sites of the first 46 institutions that received NIH-funded Clinical and Translational Science Awards (CTSAs). We reviewed all the policies we identified (through both the Google search and the CTSA institution Web site review) that provided guidelines for the distribution of F&A cost recovery, and we categorized them, post hoc, into one of four major types. We then reviewed each policy to determine whether or not it specifically addressed the issue of sharing F&A cost recovery when multiple PIs, departments, units, or centers contributed to the research. We further categorized the policies according to their specific details for cost sharing (e.g., based on investigator time, space allocation, and/or unit resources). Finally, using the same search phrases used to locate institutional policies, we conducted searches of both Google and the PubMed database to identify available literature on sharing F&A cost recovery.
On the basis of information gathered through our Internet survey of existing institutional policies, we developed an online survey to elicit information from faculty investigators and university research administrators regarding (1) their experiences with and attitudes toward sharing F&A cost recovery, (2) the types of policies, if any, their institutions had in place, and (3) their level of satisfaction with these institutional policies.
Survey sample and procedures
In early 2010, we sent invitations to participate in our survey to two groups: (1) PIs of the CTSAs and (2) members of the National Council of University Research Administrators (NCURA). The NIH CTSA program, funded by the National Center for Research Resources, was designed to create a definable academic home for clinical and translational research. We chose to survey the PIs of the 46 CTSAs (awarded from 2006 to 2009) to garner faculty input on the surveys because of the program's unique interdisciplinary emphasis, which seeks to accelerate collaboration across disciplinary boundaries, schools, and/or departments. NCURA is an international professional organization of over 7,000 members who engage in the administration of sponsored programs, including research, education, and training at colleges, universities, and teaching hospitals, and we surveyed NCURA members to gain an administrative perspective on F&A cost recovery. We sent the e-mail invitation to participate in the survey to members who subscribe to any of the following four NCURA online neighborhood listservs: Departmental Administration, Academic/Medical Center Departmental Administration, Financial Research, and Pre-Award. The members of these four listservs total approximately 1,275. We followed all NCURA neighborhood listserv rules (e.g., no commercial use).19 We sent two follow-up e-mails to the CTSA PIs and one to the NCURA neighborhood listservs.
Invitations to both groups indicated that the purpose of the survey was to gain a better understanding of not only the range of policies for sharing F&A cost recovery currently in place at universities and AHCs but also the success, as measured by user satisfaction, of these policies. In addition, the e-mail indicated that the survey was brief (requiring less than five minutes to complete), participation was voluntary, and all individual responses would remain anonymous. We did not offer incentives to survey respondents for participating. The Columbia University institutional review board approved this study as exempt.
We used Survey Monkey (http://www.surveymonkey.com) to design the online survey and collect responses. The survey included six questions regarding F&A policies and respondents' levels of satisfaction with the policies. We did not collect job title or institutional information from the CTSA group because, as each PI is a faculty member at an AHC, doing so would have compromised his or her anonymity; however, the NCURA survey included two additional questions regarding respondent's position (central or departmental administration) and institution type (predominantly undergraduate college [hereafter, “college”], multischool university with graduate students [hereafter, “university”], or university with a medical school and associated clinics [hereafter, AHC]). We designed the survey so that multiple responses from a single individual were not possible. We pilot-tested the survey with several senior faculty members of Columbia University's CTSA for possible errors, clarity, and ease of use.
Variables for analysis.
The overall goal of the statistical analyses was to correlate institutional and policy characteristics with two main response variables: (1) the respondents' views of the helpfulness of policies for sharing F&A cost recovery for supporting interdisciplinary research, and (2) the respondents' satisfaction with the policies in place (if any) at their home institutions. With respect to the first outcome, we asked respondents to indicate their level of agreement with the following statement: “Having a standard policy for F&A cost sharing helps investigators from multiple departments, units, or centers to engage in interdisciplinary research collaboration.” For all analyses, we collapsed the responses to this question into two categories: strongly agree/agree versus neutral/disagree somewhat/disagree strongly. The second outcome focused on level of satisfaction with the policy at the institution, and we also collapsed these answers into two categories: very satisfied/somewhat satisfied versus neutral/somewhat dissatisfied/very dissatisfied.
Covariates assessed included the sampling group of the respondent (CTSA versus NCURA), the existence of a standard policy at the institution (yes versus no/unsure), the type of policy in place—if applicable (one of four, determined post hoc [see Results]), and the institution's policy on allowing units to negotiate special arrangements on a case-by-case basis (yes versus no/unsure). One question asked whether F&A sharing occurred only when multiple PIs were in place or if sharing extended to situations in which any coinvestigators (PI or non-PI) were based in different departments or units. Next, a question with the descriptions of types of policies honed in on whether F&A sharing took into account the percentage of effort of the investigators involved, the location of the physical space required to complete the work, or a mix of personnel time and other resources used (e.g., space). The survey also asked whether F&A sharing arrangements were negotiated on a case-by-case basis rather than by a standard formula.
We describe responses by reporting percentages for all variables, both separately by group (CTSA or NCURA) and combined. We used risk ratios and chi-square tests to assess associations between two categorical variables, and we employed the Mantel–Haenszel approach to assess the association between two binary variables, while controlling for a multicategory predictor variable.20 We performed all analyses using STATA software (version 10.1, College Station, Texas).21
Table 1 presents the results from the Internet searches. Through the Google searches, we located 54 policies posted for AHCs, universities, and colleges as well as five institutional reports from ad hoc task forces or committees convened specifically to address the issues of sharing F&A cost recovery. In searching the Web sites of institutions awarded CTSAs, we located 19 posted policies and five institutional reports addressing the issue. Through consensus, we determined that all 19 policies fit in one of four categories or types: based on (1) the distribution of work by investigator effort, (2) the physical space used (e.g., office space, lab space, patient exam space), (3) a preset percentage formula taking into account a wide variety of factors, or (4) case-by-case basis negotiations (Table 1). For illustrative purposes, we highlight the publicly available policies that are available on the Internet for the University of Colorado at Boulder,22 the University of Kansas–Lawrence,23 the University of New Hampshire,24 and the University of Nebraska–Lincoln,25 which provide some of the most detailed models we found for sharing F&A recovery between departments/units/centers.
Looking at the combined results of both the Google searches and the CTSA institution Web site review (Table 1), we found that a slight majority of institutional policies that address the sharing of F&A cost recovery between departments/units/centers are based, either wholly (type 1) or in part (type 3), on the distribution of work (percent effort) by individual investigators (51.1%, 23/45). But, when comparing groups, we found that most of the policies uncovered through Google searches were based only on the distribution of work by investigator effort (44.8%, 13/29), whereas the majority of policies at institutions with CTSAs were based on case-by-case negotiation (75.0%, 12/16; P = .051). Although, according to their reports, all ad hoc task forces and committees identified the issue of sharing F&A cost recovery as a factor in facilitating interdisciplinary research and recommended that such a policy be established, only half (50%, 5/10) offered specific guidelines or models for establishing one (data not shown).26–30 Finally, 13.7% (10/73) of the policies we located through our Internet searches (Google, CTSA institution sites) describe the distribution and/or sharing of F&A cost recovery either as an “incentive” for investigators, centers, and units conducting the research or as “bridge funds” to stimulate additional research initiatives (data not shown).24,31–39
We uncovered very little about F&A cost recovery through our Google and PubMed database searches, and the few articles and books we did find simply indicated that recovering F&A costs was a “recognized barrier” to interdisciplinary research,5,7–9,11,12 described the role of F&A cost recovery in allocating space,40 or noted that universities are not adequately reimbursed for the indirect costs of conducting research.41 We located one proceedings article that offered a detailed description of one university's methods for assigning credit and distributing the return of F&A in order to break down barriers to interdisciplinary research.11 Although this article provides an in-depth look at one university's policy and subsequent increase in volume of proposals and research expenditures, it neither provided information on users' satisfaction with the policy nor reviewed other policies in existence. Therefore, we believe this is the first peer-reviewed publication to describe and examine in a systematic way policies for sharing F&A cost recovery and satisfaction with them.
Of the 46 possible respondents from the CTSA group, 37 (80.4%) returned surveys, and 217 respondents from the NCURA group returned surveys. Because individuals subscribed to any or all of the four listservs, we were unable to determine a denominator or response rate for the NCURA group. Table 2 gives the number (and percent) for responses to survey questions administered to all respondents (Part I) and those administered only to respondents from institutions with written policies for F&A sharing (Part II).
With respect to their primary roles of the 217 NCURA respondents, 135 (62.2%) indicated that they work in central administration, 74 (34.1%) reported that they work in departmental administration, and 8 (3.7%) did not indicate their primary role. Approximately one-fourth of the NCURA respondents (27.7%, 60/217) were from colleges, 33.6% each were from universities and AHCs (73/217, respectively), and 5.1% (11/217) did not answer.
Whereas the majority of both CTSA and NCURA respondents (80.7%, 205/254) agreed, strongly or somewhat, that a standard policy for sharing F&A cost recovery would facilitate interdisciplinary collaboration, only 35.4% (90/254) reported that their institutions had such a policy (Table 2). Although nearly two-thirds (64.6%, 164/254) indicated that their institutions did not have a policy in place or that they were unsure if one existed, 59.8% of these (98/164) confirmed that individual units within the institution are permitted to negotiate unique F&A sharing arrangements on a case-by-case basis.
Of the 90 total CTSA and NCURA respondents who indicated that their institutions had an F&A cost recovery sharing policy, 85 answered three additional questions describing their policies and their satisfaction with them (Table 2). Most of these respondents (77.6%, 66/85) reported that their policy applied to grants with any coinvestigators across different departments or schools (regardless of whether the grant involved multiple PIs), and the majority (80%, 68/85) reported being satisfied (very or somewhat) with the policy in place. Although a few respondents were either neutral (14.1%, 12/85) or somewhat dissatisfied (5.9%, 5/85) with their policy, none expressed a high level of dissatisfaction (Table 2).
Respondents from both the CTSA and NCURA groups reported four types of formulae for F&A cost recovery sharing that correspond with the four types we identified through our Internet searches (Table 2): distribution based solely on investigator effort (43.5%, 37/85), distribution based on a more complex formula that takes into account multiple factors such as personnel time and space requirements (29.4%, 25/85), calculations made on a case-by-case basis (24.7%, 21/85), and distribution based on physical space needed (2.4%, 2/85). Respondents at institutions with an F&A cost recovery sharing policy in place (yes versus no/unsure) were significantly more likely to endorse the idea that implementing a standard policy supports interdisciplinary collaboration (89% versus 76%, relative risk = 1.17; 95% confidence interval = 1.04–1.31; P = .014; data not shown). This difference remained significant (P = .012) when we adjusted for survey group (CTSA versus NCURA), as captured in Figure 1. When we compared satisfaction levels across types of policies and formulae in place, we observed no strictly statistically significant differences (P < .05). However, we did note a marginally significant relationship between satisfaction (strongly/somewhat satisfied versus neutral/somewhat dissatisfied/very dissatisfied) and whether the institution imposed a formula for F&A sharing or allowed investigators to negotiate unique arrangements: Those with formulaic policies were somewhat more satisfied (84.4%) than those pursuing case-by-case negotiations (66.7%; P = .078; data not shown).
As shown in Table 2, similar percentages of CTSA and NCURA respondents endorsed the usefulness of F&A sharing (86.5% versus 79.7%; P = .34), and the proportion in each group reporting the existence of an institutional policy did not differ significantly (27.0% versus 36.9%; P = .25). Neither the distribution of policy types nor the description of the formula differed by respondent group (P = .39 and P = .35, respectively). Finally, we explored whether the type of institution (college, university, AHC through NCURA, or AHC through CTSA) impacted agreement with the principle of sharing, policy types, and satisfaction with policies (if in place), but we found no significant differences among these variables. We did see a significant difference, however, with respect to satisfaction level among those institutions having implemented a formal policy. Among these 85 respondents from the CTSA and the NCURA groups combined, satisfaction with the existing policy was almost universal for those at universities (96.7%) and somewhat lower among AHCs contacted through CTSAs (80.0%), colleges (72.7%), and AHCs contacted through NCURA (61.9%; P = .018).
Through our searches, we found that at least 45 institutions have implemented policies for sharing F&A cost recovery between departments/units/centers, and 10 have convened ad hoc task forces or committees to discuss the issue, attesting to the perceived importance of sharing F&A cost recovery in the success of conducting interdisciplinary research. Even though results of the Google searches illustrated the wide variance of policies, implementation, and attitudes among institutions regarding shared F&A cost recovery, we were able to synthesize the results into one of four unique types (Table 1), which could be used as models by other institutions looking to adopt a formal policy.
Notable differences exist in the type of policies in place at institutions with CTSAs compared with institutions without CTSAs. Whereas only approximately half (53.7%) of the policies we discovered through our general Google searches specifically address the sharing of F&A between departments/units/centers, well over four-fifths (84.2%) of the Web sites of CTSA-funded institutions addressed this issue. This finding supports the idea that at institutions with CTSAs, the administrators make concerted efforts to establish a policy that allows for sharing F&A cost recovery for interdisciplinary research initiatives. Notably, the results from the searches of CTSA-funded institutional Web sites do not align perfectly with the results from the survey of CTSA PIs. For example, none of the Web sites from the CTSA-funded institutions reported that F&A sharing policies were based on investigator effort, whereas four CTSA PI survey respondents indicated this was the type of policy at their institution. This discrepancy is likely explained by an institution's choice not to post such documents publicly but, rather, to circulate them internally.
Although survey respondents widely agreed that a policy for sharing F&A cost recovery was important for facilitating interdisciplinary research, the majority of them reported that their institution did not have a formal policy. Notably, the support for sharing F&A to facilitate interdisciplinary research was stronger among institutions that had already implemented a formal policy with a uniform strategy for sharing cost recovery (Figure 1). Generally, those who had a policy seemed to be satisfied, regardless of the details of the policy in place.
Before administering the survey, we had anticipated that the two survey groups, CTSA and NCURA, might respond differently to the survey questions given their different roles (senior investigators versus research administrators) and the types of institutions at which they worked (college, university, or AHC). Overall, however, the responses from both groups were strikingly similar; most endorsed the idea of sharing F&A to facilitate interdisciplinary research, and most were satisfied with the policies in place. The 80% response rate among the CTSA group, consisting of busy, high-profile researchers, indicates a high level of interest in this topic for research faculty. This issue is clearly relevant for administrators at nonmedically focused institutions as well, attesting to the importance of the topic in general, regardless of the type of research conducted. Despite the high degree of interest, 8% of both CTSA and NCURA respondents were unsure if their institutions had implemented a formal policy; we found this result to be surprising because we had expected that both groups would be more knowledgeable about this topic. Finally, we note that the vast majority of survey respondents who reported having a policy in place indicated that their policy was applicable whether it was for a grant with multiple PIs (as recognized by the NIH) or a grant with a single PI and coinvestigators across multiple departments. We would have postulated that sharing F&A cost recovery would be much more commonly adopted in instances of grants awarded to multiple NIH-recognized PIs than in cases of coinvestigators from different departments collaborating.
Many respondents indicated that their institutions allowed for individual F&A sharing agreements negotiated on a case-by-case basis. The slightly lower level of satisfaction among this group compared with those with standard policies in place argues against the case-by-case approach as a primary strategy. Perhaps investigators are less satisfied with case-by-case negotiation because it may entail more time (and perhaps confrontation) in the negotiation process, slowing the progress of research.
Institutions other than those counted in Table 1 may have similar policies that were not publicly posted from November 2009 to January 2010 or that were accessible only through a secure institutional log-in. As with any survey, the results are vulnerable to nonresponse bias. For example, the response rate among NCURA members was dramatically lower than that of CTSA PIs. The PIs are more likely to have experience with interdisciplinary research initiatives as well as related institutional barriers. Furthermore, CTSA PIs received the survey from a colleague known to most of them, whereas the NCURA members received the survey cold through their listservs. Hence, the lower response rate among the NCURA members is not surprising. We cannot determine whether NCURA respondents were more or less likely than research administrators who are not NCURA members to be knowledgeable about the administrative barriers to interdisciplinary research and F&A policies at their institutions. Additionally, NCURA members who are active in the organization and subscribe to the online listservs are likely to be more informed and concerned about such administrative issues. Therefore, NCURA survey respondents may be unrepresentative of the entire NCURA membership as well as unrepresentative of the makeup of the overall population of employed research administrators. Lastly, a small degree of overlap between survey groups is possible because some of the NCURA respondents may be employed at CTSA-affiliated institutions.
Our survey results show that the vast majority of investigators and administrators are in favor of F&A cost recovery sharing policies to facilitate interdisciplinary research; this philosophy aligns with current NIH recommendations that institutions develop ways of sharing F&A cost recovery as part of the initiative to increase and facilitate interdisciplinary research. The majority of faculty and administrators reported satisfaction with the policies in place at their institutions, regardless of the type of policy. In addition, those respondents at institutions with formal policies in place were even more likely to endorse the principle of sharing, adding further evidence of user satisfaction with formalized F&A cost recovery sharing strategies. Given these findings, more institutions should explore establishing formal F&A cost recovery sharing policies for their investigators as a way to reduce barriers to interdisciplinary research. Investigators perceive them as helpful in pursuing interdisciplinary lines of inquiry and seem satisfied with policies that impose standard formulae based on salary distribution, space allocation, and/or other resources needed to conduct the research. We recommend that before implementing a policy, institutions that are considering developing an F&A cost recovery sharing policy carefully review the four models described in this report and engage in a detailed analysis of the implications on departmental and center budgets. A multiyear, retrospective analysis would both allow for an evaluation of the budgetary impact of sharing F&A cost recovery over time and permit departments and centers to prepare for impending financial changes.
The authors wish to thank the Irving Center for Clinical and Translational Research for providing support for this project, and the principal investigators of the National Institutes of Health Clinical and Translational Science Awards Consortium and subscribers to the National Council of University Research Administrators listservs for their participation in the survey.
This publication was made possible by grant number UL1 RR024156 from the National Center for Research Resources, a component of the National Institutes of Health (NIH), and NIH Roadmap for Medical Research.
The survey was approved by the Columbia University institutional review board as an exempt protocol.
The content and opinions expressed in this report are solely the responsibility of the authors and do not necessarily reflect the official views of the National Center for Research Resources, the National Institutes of Health, Columbia University, or the National Council of University Research Administrators.
1Hall KL, Feng AX, Moser RP, Stokols D, Taylor BK. Moving the science of team science forward: Collaboration and creativity. Am J Prev Med. 2008;35(2 suppl):S243–S249.
2Dougherty D, Conway PH. The “3T's” road map to transform US health care: The “how” of high-quality care. JAMA. 2008;299:2319–2321.
3Wuchty S, Jones BF, Uzzi B. The increasing dominance of teams in production of knowledge. Science. 2007;316:1036–1039.
4Dubrow G, Tranby E, Voight C, eds. Fostering Interdisciplinary Inquiry: Proceedings From a Conference. Minneapolis, Minn: University of Minnesota; 2009.
5Committee on Facilitating Interdisciplinary Research and Committee on Science Engineering and Public Policy. Facilitating Interdisciplinary Research. Washington, DC: National Academies Press; 2005.
6Maglaughlin KL, Sonnenwald DH. Factors that impact interdisciplinary natural science research collaboration in academia. Paper presented at: 10th International Conference of the International Society for Scientometrics and Informetrics; 2005; Stockholm, Sweden.
7Larson EL. Minimizing disincentives for collaborative research. Nurs Outlook. 2003;51:267–271.
9Heitkemper M, McGrath B, Killien M, et al. The role of centers in fostering interdisciplinary research. Nurs Outlook. 2008;56:115.e2–122.e2.
10O'Connell KA. Research: Barriers to interdisciplinary research. J Prof Nurs. 2001;17:153–154.
11Roberts JA, Barnhill RE. Engineering togetherness (an incentive system for interdisciplinary research). In: Proceedings of the 31st Annual American Society for Engineering Education/IEEE Computer Society Frontiers in Education Conference, Reno, Nev, October 10–13, 2001. Piscataway, NJ: IEEE; 2001:F2G-23–F22G-27.
12Bruun H, Hukkinen J, Huutoniemi K, Thompson Klein J. Promoting Interdisciplinary Research: The Case of the Academy of Finland. Helsinki, Finland: Academy of Finland; 2005.
18Kulage KM. Interdisciplinary research at an academic medical center: An administrator's journey. Natl Counc Univ Res Admin Mag. 2010;42:26–27.
20Fleiss JL, Levin B, Palik MC. Statistical Methods for Rates and Proportions. 3rd ed. New York, NY: John Wiley & Sons; 2003.
21StataCorp. Stata Statistical Software: Release 10. College Station, Tex: StataCorp LP; 2007.
25University of Nebraska–Lincoln, Sponsored Programs, Office of Research and Economic Development. FAQ: Facilities and administrative (F & A) costs on sponsored projects. http://research.unl.edu/sp1/policies.shtml
. Accessed November 23, 2010.