Since early April, the National Science Foundation (NSF)—the nation’s leading funder of research and education in science, technology, engineering, and mathematics (STEM)—has canceled more than 1,500 grants and contracts, totaling more than $1 billion in research funding. The cancelations have sent shock waves through the scientific community, raising concerns about the future of American research and innovation.
NSF annually awards roughly $8.5 billion in grants (PDF) to universities, nonprofits, and other organizations seeking to advance science and technology. NSF grants support critical projects across the country, including advancing cancer treatment and biotechnology and developing innovations in energy and artificial intelligence. Many NSF projects also seek to improve education and research opportunities for historically underrepresented groups.
This year, NSF changed its funding priorities to ensure grants don’t prioritize certain groups or individuals, as guidance deemed these projects to “not effectuate NSF priorities.” Here, we examine the application of NSF’s new priorities by analyzing where projects have been canceled and—given the Trump administration’s pushback on diversity, equity, and inclusion (DEI)—whether the presence of DEI-related terms and concepts in the titles and descriptions have led to higher rates of cancelation.
Where are canceled NSF grants located?
As of April 18, 2025, NSF had funded 40,715 active projects since 2020, totaling about $28 billion. Of those, NSF has canceled 1,574 projects since April 18, according to an open database project called Grant Watch, run by Noam Ross, executive director of the nonprofit rOpenSci, and Scott Delaney, a researcher at Harvard University. Using the Grant Watch database, we can compare canceled grants to all active NSF grants and assess how much grant money different areas of the country are losing.
The 1,574 canceled awards total about $1.1 billion, or roughly 4 percent of all active NSF projects, but certain areas of the country have been hit harder than others. Some congressional districts have had more than one-quarter of total NSF funding canceled since April 18, 2025, while others have had no NSF projects canceled. Unsurprisingly, given the administration’s ongoing fight with Harvard University, Massachusetts’s Fifth Congressional District has lost the largest amount of NSF dollars, more than $165 million, or about 40 percent of total NSF dollars in the district.
Colorado’s Second District—home to the University of Colorado Boulder and Colorado State University, which ran several projects supporting multidisciplinary and collaborative research efforts in biology and engineering—has had more than $29 million in grant funding canceled (4.3 percent of the total), the third-highest among congressional districts. And Arizona’s Fourth District, where Arizona State University is located, has had more than $26 million in canceled NSF grants.
The dollar amounts of canceled grants don’t tell the whole story—the share of total canceled NSF grant dollars better illustrates the economic impact on institutions and the communities in which those institutions are based. In California’s Forty-Eighth Congressional District (inland San Diego County), the Southern California Tribal Chairmen’s Association had a nearly $1 million grant canceled, the only NSF grant in that district. In Texas’s Fourteenth District, a $2.9 million grant at Galveston College was canceled (amounting to nearly half of all grant dollars in that district). Of the 432 congressional districts in the data, 23 had at least 20 percent of all NSF grants canceled (totaling $218 million), and 305 had fewer than 4 percent of their existing NSF grants canceled (totaling $398 million).
What determined whether a grant was canceled?
NSF hasn’t released detailed information about why some grants were canceled and others weren’t. However, given the administration’s efforts to push back on DEI and DEI-related terms and concepts, we started there. The New York Times published a list of words that federal agencies have flagged to limit or avoid, compiled from federal agency guidance and documents. We conducted a text analysis to determine which of the canceled grants included those words or phrases (including minor variations) in the detailed abstracts.
We started by grouping similar words from titles and abstracts of NSF awards into nine distinct thematic categories. For example, the diversity, equity, and inclusion category includes those three words and variations such as “underrepresented,” “inclusive,” “equitable,” and “systemic.” The large race and ethnicity category includes words like “racial,” “black,” “minority,” and “identity.”
Overall, we find that canceled contracts are far more likely to include key words in the DEI, race and ethnicity, gender and sexuality, and governance and policy categories than grants that haven’t been canceled. In fact, nearly 90 percent of all canceled projects included at least one word in the DEI category compared with about half of projects that have not been canceled, which appears to be consistent with NSF’s new stated priorities and the administration’s stated goals.
Limiting this analysis to specific keywords in award titles and abstracts that appear at least 50 times, we find that words like “diverse,” “underrepresented,” “institutional,” “historically,” and “racial” are frequently associated with canceled NSF grants.
Overall, about 94 percent of all canceled grants included at least one of the words in the New YorkTimes list. Of the remaining 6 percent of canceled awards, more than three-quarters were awarded to Harvard University, and all but one were focused on misinformation or social media. Among the group of noncanceled grants, about 67 percent included at least one of the words in the New YorkTimes list.
What does this mean for future NSF grants?
Because NSF hasn’t explained why some grants were canceled and others weren’t, this analysis only suggests there are patterns in the words and phrases included in certain projects. We find that projects focusing on improving or expanding access to science, technology, engineering, and math education and resources to underrepresented and marginalized groups were more likely to be canceled than other projects.
These cuts risk limiting the economic opportunity for certain groups of students and hurting American innovation overall. Research has shown how expanding access to education raises incomes and opportunities, especially for students in marginalized groups. America’s strong STEM education system has positioned it as a global leader in science and technology, a position that could be undermined without further resources.
How canceling these NSF projects in 2025 will affect access to research and educational opportunities for all students may not be seen for years. But it’s certain that canceling these projects will have consequences for the researchers and scientists directly engaged in the research, the schools and communities that rely on these funds, and the students and educational programs the projects sought to improve and expand.
Let’s build a future where everyone, everywhere has the opportunity and power to thrive
Urban is more determined than ever to partner with changemakers to unlock opportunities that give people across the country a fair shot at reaching their fullest potential. Invest in Urban to power this type of work.
