NSF Budget: FY22 Outcomes and FY23 Request

Congress increased the National Science Foundation’s budget by 4% to $8.8 billion for fiscal year 2022, well short of the Biden administration’s proposal for a 20% boost. The administration is aiming high again for fiscal year 2023, requesting a nearly 20% increase to $10.5 billion.

Despite its ambitiousness, this target budget trails the amounts recommended in the landmark innovation policy bills that Congress is aiming to reconcile this year: the House’s America COMPETES Act of 2022 and the Senate’s U.S. Innovation and Competition Act (USICA). The COMPETES Act proposes NSF receive $12.5 billion for fiscal year 2022 and $14.6 billion for fiscal year 2023, while USICA proposes $10.8 billion and $12.8 billion, respectively.

A significant fraction of the budget increases sought by these bills and by the administration would go toward establishing a new arm of NSF focused on use-inspired research, which the agency refers to as the Directorate for Technology, Innovation, and Partnerships (TIP). The administration seeks $880 million for the TIP Directorate in fiscal year 2023, $15 million more than it requested last year, with roughly 40% of the total coming from existing programs transferred into the directorate. Meanwhile, the COMPETES Act proposes a first-year budget of $1.4 billion while USICA proposes $1.8 billion.

The administration is also seeking across-the-board increases for NSF’s existing research and education directorates, with especially large boosts proposed for programs connected to climate change research, clean energy technology, and STEM workforce diversity initiatives. Meanwhile, NSF’s budget for major facilities construction would decline as no new projects are proposed and several in the pipeline are approaching the completion of their funding profiles.

Detailed congressional direction for fiscal year 2022 is included in an explanatory statement accompanying the appropriations legislation, as well as in a report prepared by House appropriators. Summary figures for the appropriation and the request are compiled in FYI’s Federal Science Budget Tracker .

TIP directorate

Congress broadly endorsed the TIP directorate concept in its appropriation for fiscal year 2022, though it did not specify an initial budget. NSF formally established the directorate on March 16, the day after the appropriation was signed into law, though it has not yet announced how much of the budget increase it received will be allocated to the new directorate. NSF has however estimated that in fiscal year 2021 it spent about $370 million on existing activities that will now be housed in the directorate, such as the SBIR/STTR small business R&D programs and the Innovation Corps entrepreneurial education program.

Of the new funds requested for the TIP directorate for fiscal year 2023, the administration seeks $200 million to launch up to 10 Regional Innovation Engines, which would aim to spur economic growth in parts of the U.S. that “have not fully participated in the technology boom of the past several decades.” The Engines would fund collaborations between industry, academia, state and local governments, and venture capital organizations, among others. NSF notes the Engines would be funded through ten-year awards with budgets that are “an order of magnitude greater than traditional NSF center-scale awards.”

The fiscal year 2022 appropriation encourages NSF to establish at least one Engine this year, and the report from Senate appropriators directs NSF to establish at least 20% of the Engines in EPSCoR jurisdictions , which are U.S. states and territories that receive a relatively small fraction of the agency’s budget.

Among the other new activities proposed within the TIP directorate is an Assessments for Science and Technology Investments program, initially funded at $40 million, that would evaluate the effectiveness of major federal R&D spending initiatives. The budget also includes $25 million for an Entrepreneurial Fellows program that would provide technology commercialization training to scientists and engineers, and $20 million for a Technology and Innovation Internships for Experiential Learning program that would “support internships and other experiences, training, and credentialing for diverse learners at every stage of education, from first-time job seekers to experienced workers looking for new opportunities.”

The scope of the directorate may be significantly affected if Congress reaches agreement on a compromise to the COMPETES Act and USICA, which present competing visions for its mission. Congress has just appointed a conference committee to lead the negotiations.

Other research directorates

Across NSF’s existing six research directorates, the requested increases range from about 10% over fiscal year 2021 levels for Math and Physical Sciences to around 20% each for Geosciences, Engineering, Biological Sciences, and Social Sciences.

Congress typically does not set specific budgets for these directorates, in contrast to the level of control it exerts over other science agencies, and NSF is still determining their allocations for fiscal year 2022. However, the fiscal year 2022 appropriation broadly directs NSF to “allocate no less than the fiscal year 2021 enacted levels to maintain its core research levels, including support for existing scientific research laboratories, observational networks, and other research infrastructure assets.”

The appropriation also offers specific funding levels and policy guidance for certain activities. Outcomes and proposals for selected programs are summarized below.

EPSCoR. Congress directs NSF to increase funding for its EPSCoR program by at least 8% to $215 million, and the administration requests a further 15% increase for fiscal year 2023. The budget request also reports that NSF provided just over $1 billion to EPSCoR jurisdictions in fiscal year 2021 across all agency programs, about 12% of its total budget for that year. The figure is an important benchmark as Congress is currently considering establishing a mandatory minimum fraction of the NSF budget that must go to EPSCoR jurisdictions. USICA would require that at least 20% of NSF’s total budget go to EPSCoR jurisdictions, while the COMPETES Act proposes means of building research capacity that are not limited to EPSCoR jurisdictions. (Update: The initial version of this article incorrectly reported the EPSCoR budget increase for fiscal year 2022.)

Climate and clean energy. Congress directs NSF to spend at least $900 million in fiscal year 2022 on clean energy technology activities and its contribution to the U.S. Global Change Research Program. The budget request reports that NSF spent $950 million on these efforts in fiscal year 2021 and proposes an increase to $1.4 billion for fiscal year 2023. Among the specific initiatives proposed is a National Discovery Cloud for Climate, which would “federate access to advanced compute, data, software, and networking resources from multiple sources,” such as NSF-funded advanced computing facilities and commercial cloud computing resources. NSF also seeks to launch an open science initiative focused on climate, expand support for greenhouse gas monitoring, and create regional hubs focused on “climate innovation, mitigation, adaptation, and equity.”

Emerging technologies. Congress directs NSF to spend at least $636 million on artificial intelligence research and $220 million on quantum information science research in fiscal year 2022, of which $50 million is for the research centers established pursuant to the National Quantum Initiative Act. NSF reports that it spent about $700 million on AI and $255 million on QIS through its research and education directorates in fiscal year 2021, and requests increases to $734 million and $261 million respectively for fiscal year 2023. The proposal represents a leveling off of NSF’s funding for the fields following a period of rapid expansion. NSF requests larger budget increases in percentage terms for other priority technology areas, including microelectronics, biotechnology, and advanced wireless communications, which together with AI and QIS are bundled under the label of “emerging industries.”

Among new activities planned are continued expansion of the agency’s network of AI Research Institutes and formation of a National Virtual Lab for Quantum Information Science and Engineering, described as a coordination mechanism and a means of promoting “broad participation, diversity, equity, and inclusion in QISE.” The administration has embraced the virtual lab construct in other contexts, proposing that the Department of Energy establish a National Virtual Climate Lab modeled on the National Virtual Biotechnology Lab that DOE established during the pandemic to coordinate access to resources across its network of national labs. Another planned initiative is a New Future of Semiconductors (FuSe) program that will work to improve semiconductor foundry access for NSF-funded researchers as well as to develop systems and devices that address “both near-term supply chain concerns and longer-term Post-Moore’s Law challenges.”

Meanwhile, NSF proposes to ramp down its contribution to the National Nanotechnology Initiative from $612 million to $435 million, with the steepest cut falling on efforts to promote the “responsible development” of nanotechnologies.

Diversity initiatives. The largest new STEM workforce diversity initiative proposed in the request is GRANTED (Growing Research Access for Nationally Transformative Equity and Diversity), a $50 million program that would help “emerging and underserved” research institutions build capacity in grant proposal development and research administration, with an initial focus on minority-serving institutions. NSF also seeks to ramp up several existing workforce diversity efforts, for instance proposing to double funding for the MPS Ascend postdoctoral research fellowship program to $20 million within the Math and Physical Sciences Directorate. Additional diversity initiatives are funded through NSF’s education directorate.

Research security. NSF requests $2.5 million for its recently created Research Security Strategy and Policy activity. The funds would in part support a new Research on Research Security program, with goals such as “improving the quantitative understanding of the scale and scope of research security risks” as well as “assessing the additional research security risks in an innovation system that includes more use-inspired research rather than staying well within the bounds of fundamental research.” The scope of the program will be informed by a forthcoming study conducted by the JASON advisory panel, building on a study on fundamental research security the panel completed in 2019.

International partnerships. NSF proposes to increase funding for its Office of International Science and Engineering by nearly 50% over fiscal year 2021 levels to $74 million. The new funds would help launch a Global Centers program that would support international research collaborations on “grand societal challenges,” with an initial focus on climate change and clean energy.

Research infrastructure

The fiscal year 2022 appropriation provides the $249 million NSF requested for the Major Research Equipment and Facilities Construction (MREFC) account, an $8 million increase over the amount enacted for fiscal year 2021. The funds will largely support ongoing work on the Vera C. Rubin Observatory in Chile, upgrades to particle detectors at the Large Hadron Collider in Europe, infrastructure modernization in Antarctica, and acquisition of new ocean research vessels.

For fiscal year 2023, NSF requests $187 million for MREFC, which includes a final annual installment for the research vessel project and a penultimate installment for the Rubin Observatory. The agency also notes it received supplemental funds through pandemic and hurricane recovery legislation that have helped defray increased project costs, and that it is still working to assess pandemic-driven delays across its construction portfolio.

Among currently operating facilities, NSF reports that the loss of science output due to curtailed operations amid the pandemic in most cases did not result in higher costs to the agency. However, it does report higher costs attributed to pandemic mitigation protocols at certain facilities, such as those in Antarctica.

Antarctica facilities. As part of its efforts to keep the coronavirus from reaching the continent, NSF did not deploy construction personnel to its McMurdo base on the coast of Antarctica for the construction seasons in fiscal years 2021 and 2022, and on-ice construction is now scheduled to resume in October 2022. The delay has compelled NSF to broaden its Antarctica Infrastructure Modernization for Science project to encompass pressing needs in other parts of the continent not within the initial scope. A new cost baseline for the project will be established this year. NSF also reports that crew size limitations have resulted in a delay of at least three years to a planned upgrade to the IceCube neutrino observatory at the South Pole.

Rubin Observatory. NSF estimates the pandemic has caused a nearly two-year delay in completing the Rubin Observatory and notes the authorized project cost has been increased by nearly $100 million to $571 million due to the pandemic and “new data security requirements.” With the project now nearing completion, NSF anticipates ramping up its share of funds for observatory operations from the present level of around $5 million to $20 million in fiscal year 2023, rising to a bit under $40 million in future years. The total annual operating cost of the observatory is anticipated to be about $70 million, split between NSF and the Department of Energy, with in-kind contributions from international partners.

LHC upgrades. NSF states that pandemic impacts on the LHC detector upgrades have been relatively small to date because the initial efforts were focused on design work and software development. However, the agency anticipates the impacts will soon grow due to component fabrication delays caused by factors such as the global shortage of semiconductors, and it notes CERN recently announced a one-year delay to new component installation, from January 2025 to January 2026. NSF anticipates requesting $20 million above the baseline amount previously planned for fiscal year 2024, raising its share of the total project cost to about $173 million. DOE is also funding a large share of the project, but has not yet released details of its latest budget request for the facility.

Mid-scale research infrastructure. NSF requests level funding of $76 million from the MREFC account for the Mid-scale RI-2 program, a recently created agency-wide program that supports projects with costs ranging from $20 million to $100 million. Separately, NSF requests a roughly 50% increase to $50 million through its research account for the Mid-scale RI-1 program, a similar agency-wide program that supports projects in the $6 million to $20 million range. In fiscal year 2023, NSF anticipates issuing a new solicitation for the Mid-scale RI-1 program and funding new projects proposed in response to a previous solicitation for the Mid-scale RI-2 program. To date, the latter program has funded a beamline at the Cornell High Energy Synchrotron Source, a global ocean biogeochemistry array, an electric grid research testbed, a nuclear magnetic resonance infrastructure network, and a social science data resource, each with a total cost in the range of about $30 million to $50 million.

Project planning. NSF does not request funds to begin construction of any major new facilities, though it does mention some projects in the design phase. These include a Leadership Class Computing Facility and a next generation Antarctic research vessel. It also briefly references projects recommended in the latest decadal survey for astronomy and astrophysics, such as the Extremely Large Telescopes and the next generation Very Large Array.

In recent years, Congress has expressed concerns about the state of NSF’s preparations for future major facilities. The latest appropriation registers “concern about the impact of current construction delays on NSF’s planning for the construction and development of the next generation of competitive large-scale facilities” and encourages NSF to develop a “comprehensive and prioritized list” of potential projects. On this front, last year NSF’s Math and Physical Sciences Directorate established an advisory panel to help it assess potential projects.

Education directorate

NSF provides the bulk of its funding for workforce development programs through the STEM Education Directorate, which before this year it called the Education and Human Resources Directorate. Congress increased funding for the directorate by 4% to just over $1 billion for fiscal year 2022, and NSF now requests an increase to nearly $1.4 billion, a portion of which reflects a proposed consolidation of funding for the Graduate Research Fellowship Program.

Graduate Research Fellowship Program. For fiscal year 2022, Congress directs NSF to fund the GRFP with $148 million from the education directorate and up to another $148 million from its research account, amounting to a total budget increase of up to 4%. For fiscal year 2023, NSF proposes to fund the program solely through the education directorate and raise its budget by 20%. The extra funds would enable the agency to increase the number of new fellows from about 2,000 to 2,750 and increase their stipend by $3,000 to $37,000.

Broadening participation. Repeating proposals from the last budget cycle, NSF seeks double-digit percentage increases for programs that support minority-serving institutions, such as the HBCU Undergraduate Program, the Tribal Colleges and Universities Program, and the Hispanic Serving Institutions program. For fiscal year 2022, Congress ultimately provided a 4% increase for the HBCU and HSI programs and a 6% increase for the Tribal institutions program.