President Trump’s fiscal year 2021 budget request for the National Science Foundation proposes to reduce the agency’s budget by 6% to $7.7 billion. Most of the cut would fall on the agency’s main research account, which would drop 8% to $6.2 billion. Meanwhile, the budget for its education directorate would remain about flat and its major facility construction projects would receive the funding necessary to stay on track.
It is not possible to quantify how the cut would be distributed across NSF’s six disciplinary research directorates because their budgets for the current fiscal year have not been finalized. Nevertheless, the requests for each directorate offer a sense of current priorities. In particular, only the computer science and engineering directorate would receive a higher funding level relative to fiscal year 2019, reflecting the Trump administration’s emphasis on artificial intelligence R&D across federal science agencies.
For directorate-level summary tables, consult FYI’s Federal Science Budget Tracker .
Artificial intelligence. Across directorates, NSF proposes to spend $868 million on AI research, which is almost twice what it spent in fiscal year 2019. Part of the increase would go toward establishing additional AI research institutes. NSF anticipates creating up to six AI institutes this fiscal year funded at up to $20 million each over five years. Some of these institutes will focus on applications of AI to research. Among the anticipated topics for the first set of institutes are “AI for discovery in physics” and “AI for accelerating molecular synthesis and manufacturing.”
Quantum information science. QIS also stands out as a top priority as the administration continues to implement the National Quantum Initiative Act . NSF proposes to roughly double its funding for the field to $226 million, in part to support new quantum research centers that it is establishing through its Quantum Leap Challenge Institutes program. NSF plans to stand up an initial set of centers this year and to hold a second competition in fiscal year 2021. NSF would also expand support for its Quantum Leap Big Idea, which is geared toward cross-directorate research.
Big Ideas. NSF proposes to continue nine of its “10 Big Ideas,” a framework the agency uses to focus support on crosscutting research initiatives. It does not request funding for the NSF 2026 Big Idea, through which the agency just completed a competition to identify candidate concepts for future initiatives.
Spectrum research. Among the new activities proposed in the budget is a “Spectrum Innovation Initiative” that would develop methods for more efficiently sharing radio bands in light of the increasing spectrum demands of emerging technologies. Out of an initial budget of $17 million, NSF proposes that $9 million go toward creating National Radio Dynamic Zones, which are described as “a novel mechanism for piloting, testing, and rolling out the most innovative approaches to dynamic spectrum sharing in specialized geographic regions.” Of the remaining amount, $5 million would fund a National Center for Wireless Spectrum Research that promotes cross-disciplinary research on spectrum sharing.
Other priorities. Other areas emphasized in the budget request include advanced manufacturing, microelectronics, biotechnology, cybersecurity, and coastal resilience.
An image of the sun’s surface taken by the Daniel K. Inouye Solar Telescope, which began operations last year. Construction of the facility was funded through NSF’s Major Research Equipment and Facilities Construction account.
(Image credit – NSO / AURA / NSF)
Major construction. NSF requests $230 million for the Major Research Equipment and Facilities Construction (MREFC) account, $13 million below the current enacted level. Of the total, $90 million is for upgrading infrastructure at McMurdo Station in Antarctica, $41 million is for the penultimate year of construction of the Vera C. Rubin Observatory in Chile, and $33 million is for NSF’s second year of contributions to a project to upgrade CERN’s Large Hadron Collider for higher luminosity operations.
The National Science Board recently gave the LHC upgrade project a formal green light , authorizing NSF to make awards to Columbia University and Cornell University for improvements to the ATLAS and CMS detectors, respectively. The Department of Energy’s budget request includes $78 million for associated upgrades.
No funds are specifically requested for the design of future MREFC projects, though the budget notes NSF recently funded the Texas Advanced Computing Center to conduct design work for a potential “Leadership Class Computing Facility.” The facility would be 10 times more powerful than the center’s Frontera petascale supercomputer , which began full operations last October.
Mid-scale research infrastructure. Of the remaining MREFC money, $65 million is for the Mid-scale Research Infrastructure-2 program, which will fund construction projects costing from $20 million to $100 million. The program received an initial appropriation of $65 million from Congress for fiscal year 2020. NSF anticipates making between three and six awards through the program’s first competition, which it will fund through multiple appropriations.
In parallel, NSF requests $53 million for the Mid-scale Research Infrastructure-1 program, which funds projects costing between $6 million and $20 million as well as smaller scale design studies. NSF issued its first 10 awards through the program last year and anticipates awarding between seven and nine awards in the next two-year funding cycle. Of the funds requested, $20 million is set aside for EPSCoR states .
Facility operations. The operations budgets for most facilities would drop relative to fiscal year 2019 under the budget request, though interpreting the figures is complicated by several facilities having received one-time funding for infrastructure upgrades that year. These include the Advanced LIGO Plus project, an adaptive optics upgrade to the Gemini Observatory, and a new aviation research facility at the National Center for Atmospheric Research.
Furthermore, in the wake of last year’s government shutdown , NSF notes it opted to give many research facilities a tranche of extra funding in fiscal year 2019 to help ensure “continuity of operations.” NSF explains it concluded major facilities should have “at least three months of funding obligated to execute the NSF mission across any recognized boundaries of funding discontinuity or other potential shortfalls.”
A handful of facilities would see budget increases under the request. With the Rubin Observatory nearing completion, NSF requests $5 million to support initial operations of the telescope, which will collect nearly 40 terabytes of imaging data each night as it surveys the southern sky.
NSF is responsible for half the facility’s operation costs, with its contribution projected to grow to $30 million annually by fiscal year 2024. While it was originally anticipated that international organizations would partially support operations, the budget document notes that NSF and DOE have instead decided to seek in-kind contributions that will offset operations costs.
Meanwhile, NSF also seeks to ramp up funding for operations of the Daniel K. Inouye Solar Telescope (DKIST) from $11 million to $18 million, near the projected peak annual amount. To help alleviate the burden such rising operations costs impose on directorate budgets, NSF requests $10 million for the second year of a pilot Facility Operation Transition program. A portion of these funds would support initial operations costs of DKIST, the Rubin Observatory, and the National Ecological Observatory Network.
Divestment. At the other end of the facility life cycle, a few of the proposed cuts stem from planned divestment actions. NSF expects to provide a final year of support to the National Superconducting Cyclotron Laboratory at Michigan State University before DOE takes over stewardship of the facility.
NSF also proposes to reduce its annual contributions to the Arecibo Observatory in Puerto Rico and transition its support for the Cornell High Energy Synchrotron Source (CHESS) to the Center for High-Energy X-Ray Science (CHEXS), a new subfacility.
The budget for the Education and Human Resources Directorate would drop $9 million to $931 million under the request.
NSF proposes to pare back its flagship Graduate Research Fellowship Program by $9 million to $275 million and anticipates awarding 1,600 new fellowships, 376 fewer than were awarded in fiscal year 2019. It also states the program will “continue to align awards with NSF and administration research priorities, including AI, QIS, and other industries of the future.”
Meanwhile, NSF proposes to nearly double the directorate’s contribution to the Research Traineeship program to $62 million, with a focus on AI-related occupations. However, NSF’s research directorates would end their funding for the program, which amounted to $21 million in fiscal year 2019.
NSF’s programs that support Minority Serving Institutions would see a range of cuts under the proposal, with the steepest falling on the recently created Hispanic Serving Institutions program, which would drop from $45 million to $14 million.
