US brain drain in the physical sciences due to federal funding cuts

# Investigate the Emerging U.S. Brain Drain in the Physical Sciences

## Overview

The United States is at risk of losing scientific talent because federal research cuts, grant cancellations, hiring freezes, agency workforce reductions, immigration uncertainty, and political interference have made U.S. research careers less stable. This is an established trend, over 15,000 scientists have already left according to Neil Degrasse Tyson.

> [!IMPORTANT]
> THIS INVESTIGATION SHOULD AVOID POLITICAL STATEMENTS AND STICK TO THE FACTS. THERE SHOULD BE NO MENTION OF LEFT AND RIGHT. THIS *IS* A POLICY ISSUE BUT HACK FOR LA DOES NOT EXPRESS POLITICAL VIEWS.

At the same time, China, the European Union, Canada, Singapore, and other countries are increasing research investment and actively recruiting scientists trained or employed in the United States.

The investigation should focus on the **physical sciences**, including physics, chemistry, astronomy, materials science, earth and atmospheric science, energy science, and closely related engineering fields.

The central research question is:

> Are federal science-funding reductions and policy instability causing physical scientists, graduate students, postdoctoral researchers, and international scholars to leave—or choose not to enter—the United States?

A January 2026 Nature analysis estimated that more than 7,800 research grants had been terminated or frozen, approximately 25,000 employees had left federal science agencies, and proposed science-budget reductions totaled about $32 billion. NSF awarded 25% fewer new grants in 2025 than its average during the preceding decade.

The U.S. physical-science workforce is particularly exposed because it depends heavily on international talent. NCSES reports that foreign-born workers are substantially represented among physical and related scientists, with the foreign-born proportion increasing at higher education levels.

The project should distinguish three related forms of loss:

* **Departure:** scientists currently working or studying in the United States relocate abroad.
* **Return migration:** international students and researchers return to their country of origin rather than building U.S. careers.
* **Foregone immigration:** scientists who previously would have chosen the United States instead select laboratories in China, Europe, Canada, Singapore, or elsewhere.

Evidence of an emerging brain drain is already present, but its full size is not yet known. Nature found increasing overseas job searches by U.S.-based researchers, while a separate poll found that roughly three-quarters of responding scientists were considering leaving. These surveys indicate severe risk but are self-selected and cannot establish the number who actually moved.

International competitors are deliberately exploiting the disruption. European governments and universities have created relocation funds and recruitment programs, while China continues to expand laboratories, research funding, infrastructure, and talent-recruitment initiatives.

### Tentative Hypotheses

**H1 — Funding losses are reducing U.S. physical-science career opportunities.**

Expected evidence:

* Fewer NSF and Department of Energy Office of Science awards;
* Lower award dollars and success rates;
* More terminated or delayed grants;
* Fewer postdoctoral, faculty, laboratory, and graduate-research positions;
* Reduced funding for major instruments, national laboratories, and university research centers.

**H2 — Early-career and international researchers are the first groups to leave.**

Expected evidence:

* Declining international graduate enrollment or acceptance rates;
* Fewer international researchers beginning U.S. postdoctoral appointments;
* More U.S.-trained researchers taking their next position abroad;
* Higher departure rates among researchers dependent on temporary visas;
* Growing movement after grant terminations or laboratory downsizing.

**H3 — Other countries are converting U.S. instability into recruitment gains.**

Expected evidence:

* New foreign programs explicitly recruiting U.S.-based scientists;
* Increased relocation grants and laboratory startup packages;
* Growth in scientists moving from U.S. affiliations to European, Chinese, Canadian, or Singaporean affiliations;
* Increased foreign employment among recent U.S. doctorate recipients.

**H4 — China is a primary long-term beneficiary.**

Expected evidence:

* Increasing movement of Chinese-born scientists back to China;
* More U.S.-trained researchers appearing at Chinese institutions;
* Increasing Chinese publication and citation share in physical-science fields;
* Expansion of Chinese laboratories and research spending in strategic fields;
* Reduced U.S. leadership in emerging research areas, facilities, and collaborations.

**H5 — The largest damage may be a weakened scientific pipeline rather than immediate mass departure.**

Expected evidence:

* Graduate students leaving research careers;
* Postdoctoral positions disappearing;
* Laboratories admitting fewer students;
* Reduced training-grant and fellowship availability;
* Falling numbers of researchers entering physical-science occupations;
* Delayed experiments and permanently dispersed research teams.

---

## Action Items

### 1. Define the Study Population

Include:

* Physical scientists;
* Physical-science doctoral students;
* Postdoctoral researchers;
* University faculty;
* National-laboratory scientists;
* Research engineers in energy, materials, quantum, nuclear, atmospheric, and space science.

Analyze separately:

* U.S.-born scientists;
* Naturalized citizens;
* Permanent residents;
* Temporary visa holders;
* International students;
* Researchers of Chinese origin where data permit ethical, aggregate analysis.

Do not treat nationality or Chinese ancestry as evidence of affiliation with the Chinese government.

### 2. Establish a Baseline

Use 2015–2024 as the pre-disruption baseline and compare it with 2025 onward.

Measure:

* Federal awards and obligations;
* Number of newly funded projects;
* Average award size;
* Agency employment;
* Doctorates awarded;
* Graduate and postdoctoral appointments;
* International enrollment;
* Scientific job postings;
* Publications and affiliations;
* Researcher movement between countries.

### 3. Quantify the Funding Shock

Collect monthly or annual records from NSF, DOE, NASA, NOAA, NIST, and other physical-science funders.

Calculate:

Track separately:

* New awards;
* Continuing awards;
* Terminated awards;
* Obligated dollars;
* Award recipients;
* Early-career awards;
* Graduate fellowships;
* Major facilities and instrumentation.

Use interrupted time-series analysis to test whether 2025 produced a statistically unusual break from prior trends.

### 4. Measure Observable Researcher Movement

Create a longitudinal researcher-affiliation dataset using publication metadata.

For each author:

1. Identify U.S. institutional affiliations before 2025;
2. Identify the author’s primary affiliation in later publications;
3. Flag a possible international move when the dominant affiliation changes from the United States to another country;
4. Require repeated evidence across multiple publications or years where possible;
5. Validate a sample using institutional biographies, ORCID records, laboratory pages, or public announcements.

Primary metric:

Compare physical sciences with biomedical science, social science, and pre-2025 physical-science trends.

### 5. Measure Pipeline Loss

Investigate whether fewer researchers are entering or remaining in U.S. science.

Track:

* International physical-science graduate enrollment;
* Doctorates awarded by citizenship and field;
* Postgraduation employment location;
* Definite U.S. employment commitments;
* Postdoctoral appointments;
* University hiring freezes;
* Grant-funded research vacancies;
* Visa issuance and denial trends where available.

The Survey of Earned Doctorates provides annual data on doctorate field, citizenship, postgraduation commitments, and expected employment location, making it one of the strongest sources for detecting changes among newly trained scientists.

### 6. Investigate Foreign Recruitment

Create a structured dataset of international recruitment programs containing:

* Country;
* Institution;
* Launch date;
* Target discipline;
* Eligibility;
* Funding amount;
* Salary or startup package;
* Relocation support;
* Number of positions;
* Explicit references to recruiting U.S.-based researchers.

Pay particular attention to:

* China;
* European Research Council;
* France;
* Germany;
* Austria;
* Spain;
* Canada;
* Singapore;
* Australia.

### 7. Test the Hypotheses

Recommended methods:

* Interrupted time-series analysis;
* Difference-in-differences using less-affected fields as comparison groups;
* Researcher-affiliation transition matrices;
* Survival analysis of continued U.S. affiliation;
* Change-point detection in awards and employment;
* Network analysis of international coauthorship;
* Field-level publication and citation-share trends.

Control for:

* Normal academic mobility;
* Publication-data lag;
* Pandemic-era disruptions;
* Retirement;
* Remote or dual affiliations;
* Previous long-term changes in Chinese research capacity;
* Differences between threatened cuts, enacted appropriations, frozen funds, and actual expenditures.

### 8. Produce the Final PowerPoint

Recommended 12-slide structure:

1. What the U.S. physical-science brain drain is;
2. Why physical science is strategically important;
3. Federal funding and workforce disruption;
4. Dependence on international scientific talent;
5. Evidence scientists are considering or seeking departure;
6. Observable affiliation changes;
7. Graduate-student and postdoctoral pipeline effects;
8. Countries recruiting U.S.-based scientists;
9. China’s role and research-capacity growth;
10. Fields and institutions at greatest risk;
11. Limitations and alternative explanations;
12. Conclusions and policy implications.

The presentation must clearly distinguish:

* Confirmed funding and employment losses;
* Confirmed individual relocations;
* Measured changes in aggregate researcher movement;
* Surveyed intentions;
* Suspected long-term consequences.

---

## Public APIs and Data Sources

### NSF Award Search API

Use the NSF Award Search API to retrieve awards by year, program, directorate, institution, investigator, amount, and research topic. NSF describes the API as a public source showing how its research funding is spent.

**Can test:** declines in physics, chemistry, astronomy, materials, geoscience, and instrumentation awards.

### USAspending API

Provides federal award and obligation data across agencies, recipients, accounts, and award types.

**Can test:** whether actual federal obligations to universities and laboratories declined, rather than relying only on proposed budgets.

### NIH RePORTER API

Useful mainly as a comparison dataset for biomedical and biophysical research. Its project API exposes award, investigator, organization, fiscal-year, and project information.

**Can test:** whether physical-science-adjacent fields experienced similar or different changes.

### OpenAlex API

OpenAlex provides open data on works, authors, institutions, topics, funders, affiliations, and institutional countries. Its API supports filtering works and authors by affiliation and country.

**Can test:**

* U.S.-to-foreign affiliation changes;
* Growth of Chinese and European physical-science output;
* Shifts in coauthorship networks;
* Loss of U.S. institutional participation in strategic topics.

This will likely be the most useful free API for directly approximating scientist movement.

### ORCID Public Data

Use public ORCID records where researchers have supplied employment and education histories.

**Can test:** individual career transitions and validate OpenAlex-inferred moves.

**Limitation:** records are voluntary and frequently incomplete.

### NCSES Data and Data Explorer

Use:

* Survey of Earned Doctorates;
* National Survey of College Graduates;
* Graduate Students and Postdoctorates in Science and Engineering;
* Higher Education Research and Development Survey;
* Science and Engineering Indicators.

NCSES’s 2026 indicators document the importance of foreign-born workers to the U.S. STEM and physical-science workforce.

**Can test:** changes in doctoral production, citizenship, employment plans, postdoctoral populations, and university R&D.

### Department of Energy Office of Science Data

The DOE Office of Science is the largest federal supporter of basic physical-science research in the United States.

Use its budget documents, award records, laboratory data, and USAspending transactions to investigate physics, chemistry, materials, fusion, nuclear science, and scientific-computing support.

### Grants.gov API and Data Extracts

Use funding-opportunity and award-related records to track:

* New solicitations;
* Cancelled opportunities;
* Reduced program availability;
* Changes in eligibility or research priorities.

### Crossref API

Use publication and funder metadata as a secondary source for:

* Funding acknowledgments;
* Publication output;
* DOI-level author affiliations;
* Research-field trends.

---

## Resources

* Nature’s 2025 analysis found early indications of overseas job-seeking by U.S.-based scientists.
* Nature’s January 2026 review documented major grant disruption, agency workforce losses, and historically low numbers of new NSF and NIH grants.
* Reuters documented European programs designed to attract scientists displaced or discouraged by U.S. policy.
* Scientific American reported in June 2026 that dozens of countries were actively recruiting U.S.-based scientists and that some relocations had already occurred.
* The Association of American Universities maintains an updated collection of reports and documented cases involving researchers leaving or declining to come to the United States.
* NCSES Science and Engineering Indicators should provide the baseline for the composition and international dependence of the U.S. scientific workforce.

### Central Investigative Caution

The evidence supports describing the situation as an **emerging or accelerating brain drain**, but the report should not claim that every scientist considering departure has already left. Surveys, news reports, cancelled grants, reduced hiring, and foreign recruitment establish the mechanism and serious risk. OpenAlex affiliation histories, NCSES career data, federal award APIs, and institutional records are needed to demonstrate the scale of completed migration.

The most consequential loss may not be a visible wave of established professors moving abroad. It may be a less visible collapse in the pipeline: international students choosing other countries, postdoctoral researchers accepting foreign positions, young scientists leaving research entirely, and laboratories becoming unable to train the next generation.

- If this issue requires access to 311 data, please answer the following questions:
- Do you need a one-time or ongoing dump of the data?
- Do you need subset of data (i.e. certain years) or the entire data set (approx. 4 million rows or 11 GB)?
- If a subset is needed, please define subset characteristics (i.e. date range, etc.)
- Do you need online access via an API or a download of data?