Graduate Education: Funding, Outcomes, and Questions About the PhD

Graduate education in the United States trains researchers, scholars, and professionals across every field of human inquiry. Doctoral education in particular is the training ground for university faculty and for researchers in government, industry, and nonprofit settings. The system has produced extraordinary intellectual output and has been a source of significant talent for multiple sectors. It has also been characterized by persistent dysfunctions that affect student wellbeing, completion rates, career outcomes, and the health of academic disciplines. The mental health crisis in graduate education is well-documented and has received increasing attention in recent years. Surveys of doctoral students consistently find elevated rates of anxiety and depression compared to general population benchmarks. The structural features of doctoral training contribute to these elevated rates: the power imbalance between students and advisors on whom their funding, degree completion, and career outcomes depend; the isolation of solo research projects; the uncertainty of completion timelines and career outcomes; and the culture of some disciplines that normalizes overwork and discourages acknowledgment of difficulty. Advisor relationships are the most consequential relational dynamic in doctoral education. Students with supportive, available, and invested advisors have substantially better outcomes across multiple dimensions: completion rates, time to degree, mental health, and career placement. Students with unsupportive, unavailable, or exploitative advisors are at elevated risk of dropping out, experiencing mental health problems, and having worse career outcomes. The power imbalance in these relationships, in which advisors control funding, project direction, and recommendation letters, makes them difficult for students to exit or challenge even when they are problematic. Completion rates and time to degree in doctoral programs are significant concerns. Completion rates for doctoral students who begin programs range from roughly 40 to 60 percent across fields, with substantial variation. Time to degree in humanities and social sciences has historically been very long, averaging seven or more years to completion. The combination of low completion rates and very long time to degree means that many doctoral students invest years in education and emerge either without the degree or with a degree obtained at an age where the returns on additional education are limited. Career outcomes for doctoral graduates have not matched the expectations embedded in doctoral training design. Many doctoral programs are designed as if all graduates will become tenure-track faculty at research universities, yet the proportion of doctoral graduates who achieve this outcome is a small fraction of program output, particularly in humanities and many social sciences. The mismatch between training and outcomes has led to calls for graduate program reform, including expanded career preparation for non-academic careers, which are the actual destinations for most graduates. Funding structures for doctoral education vary by field and institution. In the natural sciences, doctoral students are commonly funded through research grants, receiving stipends and tuition support in exchange for research assistance. In humanities and many social sciences, funding is primarily through teaching assistantships, departmental fellowships, and external fellowships. The adequacy of stipend levels relative to cost of living is a persistent issue, with many programs offering stipends that do not cover basic living expenses in high-cost cities where major research universities are concentrated. Graduate student unionization has grown as a response to these conditions, with unions at many major universities winning improved wages, healthcare, and workplace protections. STEM doctoral education has distinct features including the prominent role of external research funding in structuring training. Principal investigators with large grants employ doctoral students as research workers in ways that can advance students' training or can prioritize research production over student development depending on the advisor's priorities. The alignment between advisor research interests and student dissertation interests is important: students whose dissertations advance their advisors' research agendas may complete more quickly but may also be developing research programs that are not fully their own. Professional doctoral programs in fields like education, psychology, nursing, and social work occupy a distinct space in the graduate education landscape. These programs are designed to train practitioners rather than researchers, and their outcomes should be evaluated against practitioner career goals rather than faculty placement rates. The growth of professional doctorates, including the Ed.D., Psy.D., and D.N.P., reflects recognition that doctoral-level training for practice-oriented careers has different purposes than research doctorates and should be structured accordingly. Reform proposals for doctoral education include shortening time to degree through clearer expectations and milestones, improving advisor accountability through co-advising models and structured feedback, expanding career preparation beyond academic placements, increasing stipend levels, and developing more systematic approaches to mental health support. The variation in conditions across institutions and disciplines makes uniform reform difficult, but the case for change is strong enough that the major organizations in graduate education have increasingly acknowledged the need for reform and begun working toward it.