What GAO Found GAO found that about 62,500 graduate researchers in science, technology, engineering, and mathematics (STEM) and 22,000 STEM postdoctoral scholars (postdocs) received federal funding in academic year 2023, according to analysis of federal data. Number of STEM Graduate Researchers by Funding Source and Research Field, Academic Year 2023 Number of STEM Postdoctoral Scholars (Postdocs) by Funding Source and Research Field, Academic Year 2023 Federal agencies fund graduate researchers and postdocs either directly—such as through a research fellowship—or indirectly—such as through a grant to a university. For direct funding, agencies are responsible for setting compensation levels. Directly funded STEM postdocs earned a median annual income of $60,000 in academic year 2023—the most recent year for which data were available. Such data are not available for graduate researchers. Universities and other institutions are responsible for setting indirect funding compensation levels. These institutions set compensation based on applicable laws, established policies and practices, geographic market prices, and other considerations. Comprehensive data are not available on indirectly funded compensation, but GAO’s analysis of university information found median indirect compensation levels to be about $62,200 for postdocs and $36,000 for graduate researchers in academic year 2025. The most recent U.S. Bureau of Labor Statistics data show that a full-time worker with a doctoral degree—a similarly educated group compared to postdocs—earned about $118,000. Agencies collect varied compensation data for graduate researchers and postdocs, depending on whether they are directly or indirectly funded. For example, the Department of Defense collects stipend information for directly funded graduate researchers and postdocs but not for those that are indirectly funded. This variation creates data gaps. The National Science Foundation (NSF)—the principal federal statistical agency for the STEM workforce—also collects some data on graduate researcher and postdoc compensation but does not collect these data in a manner that allows for a detailed assessment of the adequacy of compensation. NSF would be in a better position to provide policymakers with a more complete picture of the financial health and stability of the U.S. STEM research workforce if it were to comprehensively identify gaps in data needed to fully assess the adequacy of compensation and to assess the feasibility of collecting such data. In 2022, Congress directed NSF to sponsor a study on graduate student funding, including the effects of different funding mechanisms on graduate student experiences and outcomes. As of March 2026, NSF had not engaged an entity to complete the required study, nor has it established a timeline to do so. Establishing a timeline would help ensure that the study is completed. In turn, completing this study would provide Congress and other policymakers with information to understand existing funding mechanisms’ effectiveness and help them determine actions needed to improve graduate researcher experiences and outcomes. Factors that influence graduate researcher and postdoc recruitment and retention include future career goals, development opportunities, and funding stability, according to 72 postdocs and graduate researchers who responded to GAO’s questionnaire. Respondents also identified challenges, including low pay and the cost-of-living in higher cost areas. Stakeholders GAO interviewed said the lack of benefits, such as family and caregiver-oriented support, is a recruitment challenge and may prompt postdocs and graduate researchers to evaluate whether to pursue a program. Why GAO Did This Study A robust STEM workforce drives innovation and economic growth and supports U.S. national security. Federal agencies have invested billions of dollars annually in STEM research, which includes investments in graduate researcher and postdoc training. These researchers may receive monetary compensation such as stipends, salaries, and wages, and fringe benefits such as vacation, sick leave, and health insurance. But compensation may be low relative to other professionals with the same level of education and experience. GAO was asked to examine federal compensation for STEM graduate researchers and postdocs. This report examines how many graduate researchers and postdocs receive compensation, the federal role in establishing such compensation, and how compensation-related factors influence recruitment and retention, among other things. GAO selected eight agencies for review—six that provided over 80 percent of federal funding to science and engineering graduate researchers in academic year 2021, and two additional agencies that either collect relevant statistical data or coordinate federal STEM initiatives. GAO also reviewed agency data, interviewed agency officials and stakeholders, and administered an online questionnaire to 72 STEM graduate researchers and postdocs, among other methods.

What GAO Found Millions of married Americans save for retirement by participating in a defined contribution plan, such as a 401(k). While most plans require spousal consent for beneficiary changes, few require it to remove funds (e.g., take a loan, withdrawal, or distribution). For example, money purchase and target benefit plans, which require spousal consent to actively remove funds (e.g., not use the plan’s default distribution option), account for less than one percent of all private sector defined contribution plans. These plans require a survivor annuity upon the participant’s death, ensuring the spouse receives regular plan payments unless the spouse previously consented to the designation of another beneficiary. The Thrift Savings Plan for federal workers generally requires spousal consent to remove funds but not for beneficiary changes. Among the married households where at least one spouse had a 401(k) or similar account, about one in 10 removed funds in 2021, according to national survey data. Those that removed funds typically took less than 10 percent of the total household retirement balance. When a defined contribution plan participant removes funds without their spouse’s consent, their spouse may experience financial and personal hardship, according to most stakeholders GAO interviewed. For example, the funds may be irreversibly gone, or it may create marital conflict. Stakeholders said that women are more likely than men to be negatively affected, in part because fewer women have their own retirement accounts. While there are no data on how often participants remove funds without their spouses’ knowledge, stakeholders said they think it is likely not common. Of the incidences stakeholders described to GAO, however, some cases reportedly involved severe consequences for the spouse, including losses of hundreds of thousands of dollars. The severity of the economic impact on the spouse depends on the total amount of funds taken and what proportion they represented of the retirement savings. Other factors impacting severity include personal circumstances such as the spouse’s income and how much time they have to make up lost savings before retirement. 401(k)-type Account Ownership in Married Couples Adding spousal consent requirements to all defined contribution plans could increase financial safeguards for some spouses but may delay processing or increase costs for plans and participants, according to most stakeholders. For example, spousal consent requirements could prevent participants from removing funds during a divorce without the knowledge of the spouse. Additional requirements could increase operating costs, which plans may pass on to participants. Stakeholders identified alternatives and modifications to spousal consent requirements that may reduce administrative burdens, including notifying spouses when a participant removes funds, or only requiring consent to remove funds above a certain threshold. They also said that some exceptions to spousal consent requirements may be warranted, such as in cases of domestic violence. Why GAO Did This Study Married individuals can spend a lifetime saving for retirement through defined contribution plans, and the federal government offers incentives to contribute to them. If a plan participant removes retirement funds from their account without their spouse’s knowledge, it can significantly reduce the future retirement income for both of them. Federal law provides protections in some retirement plans by requiring spousal consent to remove funds, but protections differ among plan types. GAO was asked to examine these issues. This report examines (1) when married participants are required to obtain spousal consent to remove funds from or designate a beneficiary in defined contribution plans, and how often fund removal occurs; (2) what stakeholders said about the potential effects on spouses when married participants take out funds without their spouse’s consent; and (3) what stakeholders cite as the trade-offs of increasing spousal protections and potential alternatives. GAO reviewed relevant federal laws and regulations. GAO also analyzed nationally representative survey data from the 2022 Survey of Consumer Finances, the most recent available. GAO interviewed federal government officials and stakeholders from eight national organizations. These organizations represented retirees or the retirement industry. GAO also interviewed representatives from four family or retirement law firms and five firms that sponsor or manage retirement plans, and three spouses. GAO selected interviewees based on research or referral from other interviewees. For more information, contact Tranchau (Kris) T. Nguyen at nguyentt@gao.gov.

What GAO Found Since fiscal year 2014, the Department of the Treasury has increased the size and frequency of its debt auctions to finance persistent government deficits and refinance existing debt. In fiscal year 2025, Treasury held 444 auctions of bills, notes, and bonds to borrow $1.9 trillion for government operations and refinance $9.1 trillion of maturing debt. Treasury issues debt on a regular and predictable schedule to minimize investor uncertainty. It also uses other strategies to help keep borrowing costs lower than they might otherwise be. New Borrowing and Refinancing of Treasury Securities, Fiscal Years 2014–2025 Treasury auctions continue to attract sufficient demand from a variety of investors. As of September 30, 2025, domestic investment funds—such as money market funds, mutual funds, and hedge funds—were the largest buyers at auctions, followed by broker-dealers and foreign investors. Treasury’s debt management practices alone cannot address important risks that could reduce investor demand for Treasury securities and raise government borrowing costs. In some cases, Congress would need to take action to address the underlying causes of these risks. Unsustainable federal debt levels could cause investors to demand higher interest rates to compensate for increased risk—adding to growing federal interest costs. Debt limit impasses increase the risk of a government default, which would diminish the perception of Treasury securities as safe assets. A potential diminished international role for the U.S. dollar would weaken demand for Treasury securities among foreign investors that hold dollars as reserves, use them for global trade, or use them for other financial transactions. Why GAO Did This Study As of February 2026, debt held by the public was over $31 trillion. The Congressional Budget Office projects that federal deficits will average over $2 trillion annually through 2036, further adding to U.S. debt. To finance federal borrowing, Treasury must sell large amounts of Treasury securities at auction. The interest rates that investors are willing to accept at these auctions determines the government’s borrowing costs. Thus, Treasury’s issuance decisions and auction results are important to monitor as Treasury seeks to borrow at the lowest cost over time. GAO was asked to review Treasury’s debt management practices. This report describes debt management challenges and assesses Treasury’s strategies to manage them, describes changes in debt composition, auctions, and investor demand from fiscal years 2014 through 2025, and describes other debt management risks facing Treasury. GAO analyzed Treasury data, reviewed Treasury documents and market analyses, and interviewed Treasury officials and market participants.

What GAO Found GAO identified three potentially transformative technologies that are trending toward maturity and may need congressional attention over the next 10 years. These technologies are: Neural implants for human augmentation. Currently, neural implants are only available to people with certain medical needs. Future implants might enable direct brain-to-brain communication, hands-free control of computers, or the rapid acquisition of new skills and abilities. General availability of neural implants could compromise users' privacy and security, depending on who can access data from such implants. In addition, differentiating between medical and augmentative uses would involve subjective value judgments and ethical questions. Policymakers could consider a variety of options, including determining whether to propose standards for the ethical development and use of neural implants or explore ways to ensure that privacy and security concerns are addressed. General purpose robots. General purpose robotics represents a fundamental shift from task-specific automation to flexible, adaptable machines capable of performing a wide range of tasks and potentially learning new tasks. Such robots could assist with a variety of tasks, including helping maintain infrastructure and assisting with disaster response. But their use could also lead to significant social impacts, such as risks associated with giving robots a level of autonomy in hazardous environments. Policymakers could consider a variety of options, including determining whether to explore a broad range of oversight, risk assessment, or control mechanisms (such as software controls requiring human confirmation) to help mitigate these concerns. Orbital debris removal technologies. There are more than 15,000 pieces of orbital debris currently tracked, with more than a million pieces that are too small to track but can still damage satellites and other spacecraft that provide important services. Technology is in development to actively remove, relocate, or repurpose large, non-tumbling debris. This could reduce the risk of a catastrophic cascade of collisions, but would not eliminate it because small or tumbling debris constitute the vast majority of dangerous debris. Additionally, further development and use of novel technologies may be hampered by possible legal difficulties posed by the Outer Space Treaty. Policymakers could consider a variety of options, including supporting targeted research to fill technological gaps or initiating legal analyses to develop solutions to legal difficulties. GAO is not making recommendations but has identified several policy considerations for the Congress and others to weigh as these technologies continue developing. Why GAO Did This Study Science and technology are constantly evolving, and there is a need for analysis of emerging trends of the future to help prepare for disruptions that may have major impacts in the lives of Americans. To address this need, GAO developed this report focused on technologies approximately 10 years on the horizon. The goal is to provide foresight into developing technologies that could have significant impacts on Americans. GAO described developments in these technologies and how they may be affected by various elements which may be useful for policymakers, such as legislative bodies, government agencies, or other groups, to consider. These elements include the five domains in the STEER framework: social impacts, technology drivers, environmental impacts, economic drivers, and the regulatory landscape. To conduct this work, GAO relied on a review of scientific literature from academic journals and position papers and held semi-structured interviews with eleven experts across the three technologies. GAO relied on the judgment of its engineers and scientists and consideration of the collected information to describe key aspects of the technological trends, including identifying technological developments, market conditions, or economies of scale that could further accelerate the maturity of these new technologies, and considerations for policymakers.