Last Updated on June 21, 2025, 9:47 am ET
Reimbursements to institutions for so-called indirect costs have been in the headlines this year. Even as ARL is supporting litigation that seeks to reverse a 15% cap on reimbursement rates by the US National Institutes of Health (NIH), we are also closely tracking an effort by US higher education organizations to develop a new funding model for support services that underpin the research enterprise. To keep our members informed about this complex issue, we interviewed Hilary Craiglow, who is the library consulting practice lead at Attain Partners and a national expert on indirect costs as they relate to research libraries.
ARL: How are library contributions to federally funded research reflected in the existing process by which institutions negotiate a rate for indirect costs?
HC: Libraries hold a unique position as one of only five uncapped components of a university’s indirect cost rate proposal. The current funding mechanism enables universities to recover expenses that they have already incurred in support of the research enterprise, including investments made in their libraries.
A typical ARL institution has an indirect cost rate between 45% and 60% (which translates to approximately 25.7% of the total NIH federal grant distribution). University rates vary because research costs differ across institutions. For example, costs in larger cities are significantly higher than those in more rural areas. Rates are determined by rigorous documentation of funds already spent by the institution, as documented in a facilities and administrative (F&A) cost proposal submitted to the federal government.
Negotiated every four years (with an option to extend up to eight years), each F&A proposal includes the following components:
- Administrative expenses, capped at a 26% recovery rate
- Facilities expenses, uncapped, which includes:
- Operations & Maintenance
- Building Depreciation
- Equipment Depreciation
- Interest
- Library
In the following example, Sunshine University has calculated each component as follows, which sums to a 55.00% Organized Research rate.
| Administrative | 26.00 | |
| Facilities | ||
| Operations & Maintenance | 13.90 | |
| Building Depreciation | 4.90 | |
| Equipment Depreciation | 4.15 | |
| Interest | 3.55 | |
| Library | 2.50 | |
| 55.00 | ||
Based on my experience, the average library component for R1 universities hovers around 2%.
Uniform Guidance (2 CFR Part 200) currently identifies two methods for calculating the library component: a simple full-time equivalent (FTE) method or a Special Study. In both methods, the university’s total library expenses—the library’s managed budget (with a few specific exceptions such as advertising, fundraising, alcohol, and salaries over a cap) and facility and administrative cross-allocations—are allocated to the direct functions of the institution. The following is a typical list of the functional categories in use:
- Organized Research
Research separately budgeted and accounted for by an institution, for example, sponsored, institutional, and clinical trials - Instruction and Departmental Research
Teaching, training, and departmental research without external reporting requirements - Patient Care
When applicable - Other Sponsored Activities
Sponsored programs and projects that are not research - Other Institutional Activities
University expenses that incur costs but are not instruction, research, other sponsored activities, or patient care
The FTE method, which uses student, faculty, and other user populations to calculate library expenses, is a less accurate method for determining library costs in support of organized research. It has been demonstrated to disproportionately allocate expenses associated with research to instruction. A Special Study is a more accurate accounting method and can better include the breadth of library support for the research enterprise.
ARL: How do institutions make decisions about the internal distribution of indirect reimbursements? What are the mechanisms by which libraries receive these funds, and do libraries generally receive the full library component of the negotiated rate?
HC: One of the chief complaints about the current indirect reimbursement process is the lack of transparency regarding how universities manage their recovered costs. In fairness, the expenses have already been paid by the university and are documented and audited in the F&A proposal, as well as the university budget. At most universities, however, it is unclear where the recovered dollars flow and if they are appropriately distributed to the cost centers based on actual expenses incurred. This can cause confusion or speculation.
I have advocated that university libraries recognize the value of their contributions to the research enterprise and utilize information about indirect costs in their budget discussions. However, libraries should consider thoughtfully whether they want to tie their budgets to a variable rate and to grant dollars that may or may not be received by their institution. For any library that relies on indirect reimbursements in its operating budget, we are now in a worst-case scenario with the attempt to cap all university indirect recovery at just 15%.
While many libraries do not directly receive indirect reimbursements, a 15% reduction in cost recovery would create a cascading effect on budgets across an institution. An R1 typically derives around 20% of its revenue from research activities, making indirect cost recovery crucial to financial stability. If indirect revenue drops, administrators will need to compensate by reducing baseline funding that supports core infrastructure, including libraries.
Along with libraries, other entities such as research administrators, campus maintenance, schools, and departments are also part of the indirect cost recovery rate. It is rare for any of these entities to receive direct indirect funds. Yet all are integral to the research ecosystem, not isolated service units.
In my view, it would be a mistake to dismiss concerns about current changes to indirect reimbursements simply because the money doesn’t flow directly to the library. Doing so positions the library as disconnected from the university’s research mission precisely when demonstrating that connection is most critical.
ARL: Much has been written about the effect that an across-the-board cap of 15% on indirect reimbursements would have on institutional budgets. What downstream effect would you expect this to have on the budgets of research libraries?
All of the attempts by US federal funding agencies to impose a flat 15% rate are currently in active litigation. Right now, all major implementation efforts are suspended due to court orders. We don’t know the outcome yet, but the cap is included in the FY26 federal budget that is currently making its way through Congress. Should this change eventually take effect, its impact on research institutions—and on science—would be substantial. Universities will experience significant financial losses to investments they have already made in the research infrastructure, including their libraries.
A university’s management of potential revenue reductions due to grant and indirect cost changes will vary, and is highly dependent on institutional conditions. Although there is no typical university or library, we can imagine, again, Sunshine University, which, for illustration purposes, receives a direct stream from indirect costs to its library budget.
| $ 400,000,000 | University Research Funding (Modified Total Direct Costs) |
| $ 200,000,000 | 50% Current University Recovery Rate |
| $ 10,000,000* | 2.50% Current Recovery Rate for Library |
| $ 60,000,000 | 15% Proposed University Recovery Rate (a 70% reduction) |
| $ 3,000,000 | .075% Recovery Rate for Library (a 70% reduction) |
| ($ 7,000,000) | Potential Difference for Library |
* This example, reflecting my experience with R1 universities, illustrates the potential impact of a 15% cap on libraries. It is realistic to estimate that an R1 university recovers $10 million annually from its library component, which, according to the 2023 ARL Statistics, represents a staggering 28% of a typical ARL library’s managed budget.
ARL: Our own advocacy around indirect models has focused on expanding the scope of allowable library expenses, from an outdated focus on facilities and collections to a broader set of research services. How do the draft models developed by the Joint Associations Group (JAG) stack up in this regard, and where could they be improved?
HC: Everyone is still digesting the JAG models announced last week. JAG is a coalition of ten associations of research institutions that have come together to “spur the development of a more efficient and transparent model for funding indirect costs on federal research grants.” The group has proposed two models described as “bookends,” leaving open the possibility of a hybrid model somewhere in between. Both models specifically recognize library expenses as part of Essential Research Support (ERS), a critical designation that validates these functions as fundamental research infrastructure.
Model #1, a flat rate (but almost certainly higher than 15%), takes into account differences by institution and research type and explicitly calls out “Research Library Materials” as a primary component of ERS. Beyond this reference, it is unclear how library-related support would be included, calculated, or recognized within the flat rate. However, with this model, certain library expenses could be directly charged to a grant.
Model #2, which is more challenging to implement, reflects actual expenses and, therefore, would require more detailed expense documentation than Model #1. In this model, most library-related project expenses, specifically including “scientific journal subscriptions, database access, and institutional repositories (physical and digital) directly supporting research,” would fall under “Research Information Services,” a component of ERS costs to be managed by the institution. Library costs, for this part of the model, would be calculated using a unit cost approach: namely, the cost per federally funded researcher.
Example calculation:
$800,000 in library costs ÷ 400 researchers = $2,000 per researcher
This approach is perhaps not ideal, but it aims to simplify all research-related costs through an understandable and auditable calculation. In Model #2, library-related expenses could also be assigned to the category of researcher-managed project costs.
Although Model #2 requires more detailed tracking and accounting for specialized research support expenses, something most universities and libraries do not currently do, it opens the possibility of including a more accurate account of modern research library activities. Model #2 makes library costs completely visible and auditable, while Model #1 hides them within a percentage. Model #2 would require institutions to justify and document their library cost allocation methodology.
JAG is seeking input and feedback while working diligently to collaborate with the government on a future model. If we move quickly, the library community can provide input to JAG, influencing the direction and development of a future cost model that recognizes the essential work of libraries in the research enterprise.
Through feedback to JAG, along with established connections to sponsoring organizations like COGR, ARL can provide insight into the scope of library support for research, identifying both possibilities and obstacles related to these recommendations. By examining the models with a view to library costs associated with systematic reviews, licensing, curation, and other services in addition to journal subscriptions and database access, ARL can help clarify where detailed library expenses would fit into the two models and ensure that library support of research is fully accounted for. By contributing to this process, I believe ARL can help build institutional resilience and position its member libraries as solution partners in the research enterprise.
ARL: If the research services that libraries provide to federally funded researchers were to move toward a direct charge model, then how difficult would it be for libraries to put the necessary processes for cost tracking in place? What’s one key to making this transition successfully?
HC: Under the current funding mechanism and in both proposed models, researchers can direct charge for certain library expenses that have a direct, measurable relationship to achieving the grant’s specific aims, rather than supporting general institutional operations. As examples, the following can be direct charged:
- Collections and licensing fees directly related to a research project
- Library staff salaries/benefits when working on grant activities
- Equipment and software specific to the grant project
- Space costs if library areas are dedicated to the grant project
Implementing direct charging requires a significant investment in understanding what can be direct charged, developing service and cost models, implementing new accounting systems with transparent and auditable expense tracking, policy development, and deep collaboration with university research and finance offices.
Yet the biggest challenge libraries face in moving toward direct charging is our culture. Our ethos may seem at odds with charging directly for the services, materials, and expertise we provide, which we hope benefits our entire community. We can address this challenge by itemizing certain expenses for grant purposes while still offering these same resources—to the extent that our capacity permits—without directly charging users who do not have federal funding support.
The key? A recognition of our essential place in the research enterprise, and the imperative to demonstrate our value in financial terms recognized by our institutions, funded researchers, and the federal government.
