The Musk-Trump-Cruz Assault on Science and the Disabling of American Tech

I’m sitting here digesting the news that the National Science Foundation (NSF) has laid off more than 10% of their workforce (168 people). I have spent the last 20 years working on projects focused on increasing the number and changing the demographics of students who are exposed to computing. Why change the number? Because the growth of tech, the fact that every field is a tech field, the fact that every field involves data and data analysis, requires a much larger number of people who have an ability to use computing and data analysis in effective ways. Why change the demographics? Because in order to have enough people to do the work, we need people from all parts of our population, and in order to build good technology we need representation from across the rich fabric of the country. My work in this area has focused on program building, interdisciplinary approaches, and adoption of new pedagogical approaches. And my work has been funded by multiple (6) grants from the National Science Foundation. In addition to funding foundational science, the NSF also supports K-12, post-secondary, and graduate level educational initiatives designed to increase workforce across the full spectrum of scientific fields. In fact, I had the opportunity to see the funding process up close and personal during a year I spent on temporary assignment to NSF as a program officer.

The NSF layoffs are concurrent with other attacks on the NSF awards process. These recent attacks began in October 2024, led by Ted Cruz and his aides via the Senate Ctte. on Commerce, Science, & Technology.¹ Even before the November election, Cruz et al. were analyzing the NSF database, looking for proposals that, in their view, “color scientific investigation and engagement projects through the lens of political ideology”. Their report takes to task the NSF for funding projects that include, as part of their activities, the recruitment of a demographically diverse pool of students. The screening carried out, based on word search for a large number of “DEI keywords”, identified projects such as the below that audaciously combine scientific research with efforts to broaden who in America “does” science:

• 2022 Frontiers of Engineering Symposia, organized by the National Academy of Engineering

• Creating an AI-Based community-wide efficient and equitable response system (focused on emergency services)

• A metamodeling machine learning framework for multiscale behavior of nano-architectured crystalline-amorphous composites

• The future of virtual teams: enhancing collaborative creativity and socio-cognitive wellbeing in video-based teams

• Identifying pathways to mitigate the negative impact of urbanization trends on croplands

• Flairs conference experience for workforce development in artificial intelligence

• Enabling the implementation of integrated and scalable smart cities

• Confinement induced structural evolution of calcium- and magnesium-carbonates in architected siliceous nanochannels

• Science teachers for rural America - a post-baccalaureate STEM teacher licensure project

One of Cruz’s goals is to destroy all educational and research efforts intended to increase the number of citizens prepared to work in science, including in computing and tech. Clearly his vision of science excludes everyone who has historically been marginalized from the scientific disciplines (women, people of color, rural people, people from lower socio-economic groups, disabled people). This seems very much in line with larger goals of the Musk-Trump administration, but in disagreement with parts of the MAGA base. In fact, we are really seeing a fight between different elements of the right wing, the Musk cohort of super capitalists against the Ted Cruz and Steve Bannon white supremacist MAGA acolytes.

There are two reasons why Musk is happy to see NSF efforts cut. First, he wants to fill the ensuing staffing void by increasing the H-1B visa program, letting the tech sector employ more non-citizens. Second, if universities are not funded to do scientific research, then private companies can fill the void by making profit from government contracts, much in the way that Musk’s SpaceX company has taken over elements of work that NASA used to do (in 2024, SpaceX received at least $3.8 billion in government contracts). Private companies may also hope to gain a greater financial foothold in academia by giving private grants to universities.

At the same time, Trump and his MAGA followers say they don’t want non-citizens coming into the country to work. Imagine my surprise to find that I'm in agreement with Trump confederate Steve Bannon about the apparent undermining of American labor by the tech sector. Bannon argues that “the apartheid state of Silicon Valley thinks we don’t need our greatest resource, which is the American citizen. They’d rather import basically indentured servants to work for a third less….”.

Even if you expand the H-1B program, we won’t have enough people. H-1B visas make it possible for companies in the US to recruit and hire skilled workers from other countries. New regular H-1B visas are limited to 65,000 per year, with an additional 20,000 under the “master’s cap” for advanced degree holders. Steve Bannon says “I want American citizens to get a shot at the brass ring”.² He alleges (along with others³) that the H-1B visa program is just a mechanism to put non-citizens into tech jobs at a lower salary than would be paid to citizens. Bannon claims that the Silicon Valley titans want to drive down labor costs, so they’ll push for H-1B visas, justifying it by claiming there simply aren’t enough Americans studying science & tech. But it is other parts of the Trumpian landscape, like Cruz, that are trying to ensure we don’t have the necessary home-grown talent.

Is it true that we don’t have enough citizens to fill current and future tech positions? In the period since the 2008 financial crisis, we have seen increased student interest in computing subjects. If we focus on Computer Science degrees⁴, we’ve seen an increase from 0.45% of all degrees in 2009 to 1.96% in 2023. But we have to be clear that it was still only 37,228 degrees in 2023 out of 1,894,365 total degrees (consider that 75,452 degrees were earned in Biology)! So, yes, we really don’t have enough people to fill tech jobs!⁵ To increase the number of people going into computing, it makes sense to look at who is NOT coming into the field at a sufficiently high rate. Taking a deep dive into 2023 data⁶

• Women – 0.67% of women’s degrees were earned in CS (7565 degrees), compared to 3.8% of men’s degrees (29663 degrees)

• Across race & ethnicity categories, 1.19% of degrees earned by African American/Black students (2216 degrees), 1.37% of degrees earned by Hispanic students (4597 degrees), 1.58% of degrees earned by White students (16754 degrees), 6% of degrees earned by Asian Pacific Islander students (10245 degrees)

Even though White students earn CS degrees at a relatively low rate as a percentage of total degrees, their comparatively high absolute number, along with the high number of API students, shapes the overall demographics of the computer science classroom. In computing education, the experience of an individual student is significantly impacted by the demographic composition of their classroom, and by the pedagogic practices and attitudes of all instructional staff (faculty, teaching assistants, and tutors). There is evidence-based work from a number of CS education efforts, identifying classroom structures and environments that facilitate greater recruitment and retention of students from the categories that are most underrepresented. Foundational education research shows that when we change curriculum and pedagogy with a goal of recruiting and retaining a “missing” population, all student groups benefit. So why is there now a wholesale attack on these educational activities?

The Cruz report, which presaged today’s anti-DEI attack on science education, highlights the conundrum faced by the country. We need more people who can take on roles in the tech sector, we aren’t producing enough, and we certainly need to draw more from multiple sectors of the population. Steve Bannon, ironically, is worried that Silicon Valley is creating apartheid by bringing in an indentured servant class of H-1B visas, but he doesn’t address the very real apartheid situation Cruz et al. will create by eliminating all programs designed to broaden the population that is prepared to come into the computing and tech sector. By many measures, the US has already lost any lead we had in scientific and technological research and development. Concern about this situation is clearly not a partisan issue. Heather Wilson, a Republican who represented New Mexico in the House of Representatives from 1998-2009, and who is now the President of the University of Texas-El Paso, testified before the House Science, Space, and Technology Ctte., citing the National Science Board’s Science and Engineering Indicators, reporting that “China has surpassed us in talent production, research publications, patents, and knowledge and technology intensive manufacturing”. Cruz and the Musk-Trump administration have taken actions to severely undermine the National Science Foundation and terminate funding that supports broadening the domestic populations recruited into and retained in science disciplines. Clearly their efforts are directly counter to the larger interests of the United States.

1. See D.E.I – Division. Extremism. Ideology. How the Biden-Harris NSF Politicized Science. This report is presented in an authoritative fashion but can serve as a textbook example of how data is misrepresented in order to generate an emotional response far beyond that warranted by the actual underlying information. One example is the comparison of government spending on R&D among the top five OECD countries. The report shows this data in raw numbers, with the US at the top of expenditures. It ignores the fact that the US is the largest country of the five shown, and actually spends the least per capita on R&D of the countries listed. Another example is the frequent double counting of proposals in multiple categories, coupled with problematic charts, giving the impression that the report is discussing a larger number of grants with larger dollar values than it actually covers.

2. Matter of Opinion podcast, Hosted by Ross Douthat, see Opinion | Steve Bannon on ‘Broligarchs’ vs. Populism - The New York Times

3. See, for example, Farah Stockman, Musk’s Misinformation About Tech Visas, NY Times op-ed, Jan 4, 2025

4. Using Federal CIP code 11 (Classification of Instructional Programs), the umbrella code for Computer and Information Sciences and Support Services. Computer Science, per se, is reported using the CIP subcode 11.0701.

5. Bureau of Labor Statistics indicate an increase of over 300,000 software developer jobs between 2023 and 2033. That number does not account for the need for computing personnel in an array of other job categories across the full spectrum of economic sectors.

6. 2023 is the most recent year posted to the National Center for Science and Engineering Statistics