When a university wins an NIH grant, the award has two parts. Direct costs pay for the science itself: a postdoc's salary, sequencer time, reagents. Indirect costs — officially 'facilities and administrative' costs — reimburse the institution for everything that makes the research possible but attaches to no single grant: electricity, hazardous-waste disposal, animal facilities, IT security, compliance staff, building depreciation. On February 7, 2025, NIH announced it would standardize reimbursement for new awards at 15 percent of direct costs, a sharp drop from the rates major research universities negotiate — often 50 to 60 percent or more. NIH estimated the change would have cut roughly $4 billion in annual reimbursement; within days, 22 state attorneys general sued, and on April 4, 2025, a federal judge in Massachusetts issued a preliminary injunction blocking the cap nationwide while litigation continued.
This is an explainer on research-funding mechanics, not legal or financial advice.
Where do indirect cost rates come from?
They are not invented by universities on the honor system. Each institution negotiates its rate with the federal government — for most agencies, with a DHHS cost-allocations office — through an audit-based process governed by Uniform Guidance, the federal rules at 2 CFR 200 that define which institutional expenses qualify. The university documents allowable facilities and administrative costs, divides by its base of direct research activity, and defends the arithmetic line by line. Negotiated rates vary widely because campuses vary: a coastal urban campus with expensive buildings and heavy compliance burdens carries higher indirect costs than a rural campus with newer, cheaper infrastructure. The weighted average rate across NIH awards has historically run near the high-20s in percent, lower than the headline negotiated rates because not all awards carry full rates.
What do the dollars actually buy?
The visible beneficiaries are unglamorous. Compliance offices that handle human-subjects protections, biosafety and conflict-of-interest review exist because federal rules require them. Animal facilities, freezers running around the clock, cybersecurity for controlled data, radiation-safety programs — none of these appears on a line item for any individual grant, yet all are prerequisites. Underfunding them does not make research cheaper; it pushes costs onto university operating budgets and tuition, or slows the research. That is the core argument universities made against the 15 percent cap: the rate is not overhead on science but infrastructure for it.
Related stories: Why a $50 Billion Endowment Still Says It Cannot Afford Things · How an NSF Grant Actually Gets Made, Step by Step.
How strong is the evidence on both sides?
The dispute is partly about accounting philosophy, but the underlying facts are well documented. Federal audits by Inspectors General have repeatedly examined negotiated rates and found the cost-accounting process detailed, though critics — including some economists and the federal commission that reviewed research funding under the 2017 CURE Act — have noted that cross-subsidies and weak incentives to control costs are real risks in any reimbursement model. The $4 billion estimate comes from NIH's own announcement. What no side disputes: a uniform 15 percent would have redistributed pain unevenly, hitting research-intensive institutions and, through them, the laboratories of grant-holding faculty fastest. The legal question in the injunction — whether NIH could change a system governed by negotiated agreements by announcement rather than rulemaking — remained open as litigation continued into late 2025.
Why does the fight feel sudden if the system is 50 years old?
Because the mechanics were obscure until they mattered. Cost reimbursement dates to Office of Management and Budget circulars from the late 1950s through 1960s, consolidated into the modern Uniform Guidance in 2013-2014, and for decades it ran as technocratic accounting — negotiated, audited, uncontroversial. Two things changed. Research budgets grew large enough that the indirect stream became a visible line: NIH's overall budget stood near $47-48 billion in fiscal 2024, and roughly a quarter to a third of it moved through the indirect channel. And in a 2021 case, the Supreme Court ruled that a Johns Hopkins researcher injured in a lab accident could sue under state law even though the rate system existed, drawing attention to how the government formally 'pays' these costs. By 2025 the rate had become a symbol in a broader argument about what federal research support is for — which is why an accounting rule produced injunctions within weeks.
Who pays when reimbursement falls?
Universities, in the first instance: most research institutions already subsidize research because negotiated rates recover less than full audited costs. After the cap was blocked, several universities had already made contingency planning moves — hiring slowdowns and project pauses were reported in early 2025. The secondary incidence lands on the research itself: laboratory space, staff positions and shared instrumentation funded by reimbursement. Faculty with multi-million-dollar grants are the least insulated group in this picture — their grants' productivity depends on the infrastructure the rate funds. The arithmetic most institutions fear is straightforward: at a 55 percent negotiated rate, a department absorbing the difference between 55 and 15 percent must find four dollars of institutional money for every ten dollars of direct research — money that otherwise funds hiring, space, or the institution's own research investments.
What should students and early-career researchers watch?
Three signals. Whether the litigation settles the legality of unilateral rate changes, which determines how a future cap could be imposed. Whether Congress legislates a rate instead, as several proposals in 2025 suggested it might. And whether institutions diversify support away from the federal reimbursement model. For anyone choosing a PhD program or postdoc, the practical guide is blunt: laboratory stability now depends on institutional finance as much as on grant style, and the indirect-cost line is where that dependence shows.
