nsf.gov - NCSES U.S. Doctorates in the 20th Century - US National Science Foundation (NSF) text-only page produced automatically by LIFT Text
Transcoder Skip all navigation and go to page contentSkip top navigation and go to page navigation
National Science Foundation National Center for Science and Engineering Statistics
SRS Home » U.S. Doctorates in the 20th Century » Chapter 4

Chapter 4. Path to the Doctorate



  • Attendance at 2-Year Colleges. About 9 percent of all Ph.D.s who received doctorates in 1975–84 had attended a 2-year college. By the late 1990s this proportion was 8 percent, perhaps because of an increase in the number of foreign nationals, mostly men, seeking doctorates from U.S. institutions. In the 1995–99 cohort of doctorate recipients, women were more likely than men to have attended 2-year colleges. The opposite was true in the 1970s.

  • Baccalaureates, Master's Degrees, and Professional Degrees. Most Ph.D.s earned both a baccalaureate and a master's degree before the doctorate, and most of those degrees were earned in the United States.

    As the number of foreign nationals earning doctorates in the United States increased, so did the numbers of doctoral students holding degrees from foreign institutions. By 1995–99, 28 percent of Ph.D.s held baccalaureates from foreign institutions.

  • Fields of Baccalaureate, Master's Degree, and Doctorate. Between 1960 and 1999, two in three Ph.D.s earned a doctorate in the same major field as their baccalaureate, and four in five stayed in the same field as their master's degree.

  • Financial Support. About 57 percent of Ph.D.s graduating in 1980–81 and about 58 percent of those graduating in 1996–97 received most of their financial support from research and teaching assistantships, fellowships, traineeships, internships, and dissertation grants provided by federal agencies, associations, academic institutions, and other organizations.

  • Activities in Year before Receiving Doctorate. The proportion of doctoral students employed full time in the year before graduation decreased from 44 to 35 percent between 1970–74 and 1995–99, whereas part-time employment increased from 5 to 12 percent.

    As the numbers of foreign Ph.D.s grew, assistantships became more common and fellowships less common in the year before graduation.

  • Indebtedness. In 1999, for the first time, more than 50 percent of graduating Ph.D.s reported having debt related to their undergraduate and graduate education. At the end of the century, the percentage of Ph.D.s who owed more than $20,000 in education-related debt at graduation was about 20 percent, compared with less than 7 percent in the late 1980s.

  • Time to Doctorate. Between 1920–24 and 1995–99 the median total time between receipt of the baccalaureate and receipt of the doctorate (TTD) rose from 7 to almost 11 years. The gap between men's and women's TTD narrowed. Men's TTD increased 2 years between 1960–64 and 1995–99, to 10 years, whereas women's TTD remained at about 12 years.


For students whose ultimate goal was to earn the doctorate, opportunities and options expanded in the second half of the 20th century. Many doctorate recipients still went directly from secondary schools to full-time study at 4-year colleges or universities for the baccalaureate, then immediately to universities with long-established graduate programs for full-time doctoral study. That pattern began to change, however, in the years following World War II. Some of the factors that influenced the path to the doctorate were structural changes in U.S. higher education, draft deferments during the Vietnam War, more postbaccalaureate options, consequences of the changed mix of U.S. graduate students, and growing internationalization of U.S. doctoral education. This chapter reviews the undergraduate and graduate experiences of Ph.D.s.

Attendance at 2-year Colleges

The rapid expansion of 2-year colleges during the second half of the 20th century brought opportunity for higher education closer to most high school graduates. Low tuition costs and flexible schedules, designed to accommodate both part-time and full-time students, drew large numbers of students. By 1972 these institutions were enrolling the majority of first-time college freshmen, and they did not relinquish that lead until 1986 (USED/NCES 2002, table 182). In fall 1999, 45 percent of first-time college freshmen were attending 2-year colleges.

About 9 percent of all Ph.D.s who graduated in the late 1970s and early 1980s reported that they had attended 2-year colleges, a slightly higher percentage than in the early 1970s (figure 4-1 Figure.). The percentage of S&E Ph.D.s reporting attendance at 2-year colleges decreased in the late 1980s and the 1990s, whereas the percentage of non-S&E Ph.D.s increased except for a brief period during the 1980s. In 1995–99 non-S&E Ph.D.s were almost twice as likely as their colleagues in S&E fields to have attended 2-year colleges (12 percent compared with 6 percent, respectively); for Ph.D.s in all fields the figure was 8 percent. By major field, percentages ranged from a high of 15 percent in education to a low of 4 percent in engineering (figure 4-2 Figure.).

Trends in the proportions of Ph.D.s reporting attendance at 2-year colleges were affected by the growing presence of foreign nationals receiving doctorates from U.S. institutions. The surge in male foreign students during the 1980s and 1990s was undoubtedly a factor in the percentage of male Ph.D.s who attended a 2-year college (figure 4-3 Figure.). By 1995–99 attendance at a 2-year college among those who received doctorates was more common for women than for men (9 percent compared with 7 percent, respectively), whereas the opposite was true in the early 1970s.

More than 11 percent of all U.S. citizens awarded doctorates in 1995–99 had attended 2-year colleges, up from about 10 percent in the late 1970s (figure 4-4 Figure.). Reports of attendance at 2-year colleges increased for American Indians/Alaskan Natives, blacks, and whites. These colleges played an especially prominent role in the education of American Indians/Alaskan Natives and Hispanics who eventually earned doctorates. More than one-fifth (22 percent) of American Indians/Alaskan Natives and about one-sixth (16 percent) of Hispanics who received doctorates in 1995–99 reported attendance at 2-year colleges (for Hispanics who received doctorates in 1975–79, the proportion was even higher, 21 percent). The percentages for blacks and whites were lower; the figure was lowest for Asians/Pacific Islanders (about 6 percent of those graduating in 1995–99, down from 8 percent in 1975–79).

Baccalaureates, Master's Degrees, and Professional Degrees

Throughout the 20th century, most Ph.D.s earned both a baccalaureate and a master's degree before earning the doctorate, and most of those degrees were conferred in the United States. As the number of foreign nationals earning U.S. doctorates increased, so did the number of U.S. doctoral students holding baccalaureates and master's degrees from foreign institutions. Some U.S. citizens, mostly naturalized citizens, also received baccalaureates and master's degrees from foreign institutions. More than one-fourth (28 percent) of Ph.D.s graduating in 1995–99, mainly foreign nationals, held baccalaureates from institutions outside the United States (figure 4-5 Figure.). About 90 percent of temporary resident Ph.D.s had earned their baccalaureates at foreign institutions.

Because non-U.S. citizens tend to be concentrated in S&E fields, Ph.D.s who hold baccalaureates from foreign institutions are generally more common in S&E fields than in non-S&E fields. In 1995–99, 35 percent of S&E Ph.D.s had foreign baccalaureates, compared with 15 percent of non-S&E Ph.D.s (figure 4-6 Figure.).

About 90 percent of Ph.D.s graduating in 1995–99 earned a master's degree before the doctorate; however, in many institutions a master's degree is not a prerequisite. Requirements for a master's degree also vary by field of study. For example, in 1995–99, 38 percent of Ph.D.s in chemistry, a specialty within the major field of physical sciences, and 37 percent of those in biological sciences did not hold master's degrees, compared with less than 1 percent of those in education (figure 4-7 Figure.).

Some doctorate recipients participate in programs that lead to dual degrees—the Ph.D. and a professional degree, such as the M.D. Others obtain professional degrees before receiving the doctorate. The percentage of Ph.D.s who earned professional degrees either before or concurrently with the doctorate, although remaining small, doubled from 1 percent of Ph.D.s in 1960–64 to 2 percent in 1995–99.1

Fields of Baccalaureate, Master's Degree, and Doctorate

A retrospective view of the educational background of Ph.D.s provides insights into differing traditions across fields of study. The top fields at the doctoral level (in terms of number of Ph.D.s) are not the top baccalaureate and master's degree fields of Ph.D.s. Historically, humanities has been the leading baccalaureate field of Ph.D.s, whereas education has been the leading doctoral field. Although two in three Ph.D.s earn their baccalaureate and doctorate in the same field, no doctoral field is built solely on baccalaureates from that same field.

Field switching between master's degree and doctoral fields occurs less often than it does between baccalaureate and doctoral fields. About four in five Ph.D.s with master's degrees stay in the same field for the doctorate.

Top Fields

The top fields of study of doctorate recipients at each degree level were the same at the end of the century as they had been in the early 1960s. Humanities has been the largest baccalaureate field of Ph.D.s: 20 percent of the 1960–64 cohort and 16 percent of the 1995–99 cohort held bachelor's degrees in humanities (figures 4-8 Figure., 4-9 Figure.). The top three master's degree fields of Ph.D.s in both 1960–64 and 1995–99 were humanities, education, and engineering.

Education was the top doctoral field of Ph.D.s in 1960–64 and remained the top field in 1995–99, accounting for about 16 percent of doctorates in both periods. Other fields that accounted for more than one-tenth of doctorates in both periods were humanities, biological sciences, and engineering. The field of social sciences accounted for nearly one-tenth (10 percent) of doctorates in both periods.

Field Switching

Even though some fields are larger at the baccalaureate and master's degree levels and others are larger at the doctorate level of study, in every year since 1960, about two-thirds of Ph.D.s earned their bachelor's degree and doctorate in the same major field (figure 4-10 Figure.). Even less field switching occurred between the master's degree and the doctorate.

Reasons for field switching may include the emergence of new fields, such as computer sciences, and the relation of a baccalaureate to a doctoral field. In the cases of education and professional and other fields, the doctorate adds competencies to complement degrees in subject-area fields. In 1995–99 more than one-half of the doctorates in computer sciences, education, and professional and other fields were awarded to individuals whose bachelor's degrees were not in those fields (figure 4-11 Figure.). The pattern was similar in the 1960–64 cohort.

Men switched fields between the baccalaureate and the doctorate less often than women did. The amount of field switching by men and women changed little between the early 1960s and the late 1990s (figure 4-12 Figure.). In each of the three citizenship groups, however, changes did occur. In the 1960–64 Ph.D. cohort, temporary residents were more likely than U.S. citizens and permanent residents to have switched fields between the bachelor's degree and the doctorate, but by 1995–99, temporary residents exhibited the least field switching.

Among U.S. citizens, Asians/Pacific Islanders switched fields the least and blacks the most; about 70 percent of Asians/Pacific Islanders in the 1995–99 Ph.D. cohort earned their bachelor's degree and doctorate in the same field, compared with 56 percent of blacks (figure 4-13 Figure.). Corresponding figures for the other U.S. racial/ethnic groups ranged from 63 to 65 percent. There was more field switching among Asians/Pacific Islanders, Hispanics, and whites in 1975–79 than in 1995–99.

Top Baccalaureate Fields Associated with Major Doctoral Fields

The top three baccalaureate fields associated with each major doctoral field are shown in table 4-1 Table.. The first-ranked baccalaureate field associated with a doctoral field is consistently that same field. The baccalaureate field ranked second is usually one closely related to the doctoral field. These patterns have changed little since 1960–64. Among the few changes in 1995–99 are the appearance of biological sciences in the lists of top three baccalaureate fields for the field of chemistry and the field of earth, atmospheric, and ocean sciences.

Although there is less field switching between the master's degree and the doctorate, engineering provides a good illustration of the kind of switching that can take place during graduate studies. Engineering master's degrees were common among S&E Ph.D.s in the 1995–99 cohort. In addition to being the primary master's degree field of engineering Ph.D.s (91 percent), engineering was the second largest master's degree field of Ph.D.s in the fields of physics and astronomy (6 percent); chemistry (6 percent); earth, atmospheric, and ocean sciences (6 percent); mathematics (3 percent); and computer sciences (12 percent).

Financial Support

The availability of financial support influences students' decisions about whether to pursue degrees beyond the baccalaureate. The kind of support available to students also affects the time it takes to complete additional degrees. Most doctoral students rely on several kinds of support for their graduate education. A study of S&E Ph.D.s who graduated in 1995 found that doctorate recipients used an average of 2.5 modes of support during graduate school (NSF 2000). More than 40 percent of these S&E Ph.D.s relied on three or more different sources of support, whereas 16 percent received support from only one source. The number of sources and the kinds of support used by these Ph.D.s varied by field of doctorate, sex, citizenship status, and race/ethnicity. Research assistantships, teaching assistantships, and personal resources mixed in a variety of ways to form the most common combinations of support for S&E Ph.D.s graduating in 1995.

Source-of-support questions in the SED have been changed several times, resulting in multiple code schemes that prevent the use of all source-of-support data in examining long-term trends. Information about the primary source of support, however, comes from a separate item and is coded consistently in the database, allowing a limited analysis of primary sources over time (because of survey changes, this analysis must be restricted to the 1980–81, 1996–97, and 1998–99 Ph.D. cohorts 2).

The majority of Ph.D.s receive their primary support from federal agencies, associations, academic institutions, and other organizations through such mechanisms as research and teaching assistantships, fellowships, traineeships, internships, and dissertation grants. The number of students supported by each of these mechanisms has varied over the years, but the cumulative percentage of students supported by all such mechanisms has remained close to 60 percent among the cohorts examined (57 percent in 1980–81, 58 percent in 1996–97, 61 percent in 1998–99). Another third or more of doctorate recipients in 1980–81 (38 percent), 1996–97 (36 percent), and 1998–99 (32 percent) relied on personal resources—loans, support from family members, and their own earnings and savings—as the primary source of support for their graduate education, but with variation by field. Personal resources were the principal mode of support for more about half (53 percent) of Ph.D.s in non-S&E fields compared with about one-fifth (2 percent) of those in S&E fields. The remaining percentage of doctorate recipients in each of the cohorts discussed above relied on other sources of support, such as state and foreign support (5 percent in 1980–81, 6 percent in 1996–97, 7 percent in 1998–99).

Activities in Year Before Receiving Doctorate

The activities of doctoral students in the year before graduation are consistent with the patterns of support reported for doctoral study, which show heavy reliance on personal resources and such mechanisms as assistantships and fellowships. Full-time employment in the year before the doctorate was received declined substantially after the early 1970s, when 44 percent of Ph.D.s were employed full time, to 35 percent in 1995–99. Part-time employment, in contrast, more than doubled between the 1970–74 and 1995–99 periods, from 5 to 12 percent (table 4-2 Table.).

Reliance on assistantships in the year before receiving the doctorate became more common and reliance on fellowships less common between the early 1960s and late 1990s. The shift may reflect the increased proportions of foreign nationals among Ph.D.s. Foreign nationals are not eligible for many of the fellowships available to U.S. citizens.


Information on the debt status of Ph.D.s at graduation has been collected since the late 1980s. In 1999, for the first time, about half of graduating Ph.D.s reported having debt stemming from their undergraduate and graduate educations. The proportion of non-S&E Ph.D.s reporting indebtedness increased from 44 percent in the late 1980s to 52 percent in 1999. Among S&E Ph.D.s, the proportion reporting debt fluctuated, but the percentage was about the same (49 percent) in 1999 as it had been 10 years earlier.

Not only did more Ph.D.s have debt at the close of the century, they also had larger amounts of debt (table 4-3 Table.). Although there were increases during the 1990s in the proportion in all debt-level categories above $10,000, the greatest increase was in the category for debt above $30,000, which rose from 2 to 12 percent. About one-fifth of all Ph.D.s in 1998–99 owed more than $20,000, whereas about 7 percent owed that amount in 1988–89.

The pattern of rising debt was most notable in non-S&E fields. The percentage of non-S&E Ph.D.s with education-related debt above $20,000 more than quadrupled between 1988–89 and 1998–99, from about 5 to 22 percent. During the same time interval, the percentage of S&E Ph.D.s with debt above $20,000 more than doubled, from 7 to 18 percent.

Time to Doctorate

Financial support through government programs and academic institutions, reliance on personal resources and loans, the extent of family responsibilities and dependents, and employment options during graduate study or between degrees all can influence the time it takes to attain the doctorate. Some of these factors are matters of personal choice or necessity. Others are related to characteristics of doctoral programs in specific fields and in particular institutions or to eligibility requirements for certain types of financial support funded by the federal government. Whether or not a master's degree is required can also affect the pace of graduate study. In addition, external events, such as wars, can postpone or interrupt a person's education, thereby increasing the time it takes to earn the doctorate.

Over the course of the 20th century, there was a gradual increase in the time it took students to earn their doctorates. This upward trend is evident in all three ways that time can be measured from available data: (1) total time to doctorate (TTD), the total elapsed calendar time between receipt of the baccalaureate and receipt of the doctorate, including time not enrolled in school; (2) registered time to doctorate (RTD), the time in attendance at all colleges and universities between receipt of the baccalaureate and receipt of the doctorate, including enrollment not related to the doctoral program; and (3) postbaccalaureate time to doctorate (PTD), total elapsed calendar years between the first postbaccalaureate attendance at the institution that awarded the doctorate and receipt of the doctorate.3 PTD includes time spent in a master's degree program if these studies were at the same institution that granted the doctorate.

Field of Doctorate

Between 1920–24 and 1995–99 the median overall TTD rose from 7 years to 11 years (figure 4-14 Figure.). The computation of RTD requires additional data elements collected only in the Survey of Earned Doctorates, so RTD is available only for the last four decades of the century. Between the periods 1960–64 and 1995–99, the median overall RTD grew from 5 years to 7 years (figure 4-15 Figure.). The fact that TTD rose more sharply than RTD in recent decades means that Ph.D.s were spending more time out of school before completing their doctorates. Both TTD and RTD were substantially shorter for S&E Ph.D.s than for non-S&E Ph.D.s. The increases for both measures were greater in non-S&E than in S&E fields.

The third measure, PTD, can be computed only for doctorate recipients who graduated in 1993–99. Although data from this measure cover just 7 years, it appears that the time spent in graduate work at the doctorate-granting institution (median PTD of 6 years), is changing little despite changes in the other measures. Moreover, the results for 1993–99 show little difference in PTD among the major fields.

For the 1995–99 cohort, the median PTD was the same (6 years) for S&E fields and non-S&E fields in the aggregate (table 4-4 Table.). There was some variation by subfield, but only humanities Ph.D.s, at a median of 7 years, exceeded the overall median PTD.

There was greater variation by field in TTD and RTD. Historically, chemistry Ph.D.s have had the shortest times to completion of the doctorate. In 1995–99, their median TTD was 7 years and their median RTD was 6 years. Master's degrees were least common among chemistry Ph.D.s, so their shorter times to completion may be partly explained by a greater tendency to move directly into a doctoral program soon after receiving the baccalaureate. In contrast, the field of education typically requires students to have several years of work experience in elementary or secondary schools before receiving the doctorate, with the result that the time between baccalaureate and doctorate is relatively long compared with other fields. Doctorate recipients in education have taken the most elapsed time to complete their degrees: a median TTD of 20 years in 1995–99.

Demographic Groups

Many factors, including domestic and international conditions, influence the time it takes graduate students to earn the doctorate. Furthermore, fields differ in terms of the availability of financial support for full-time graduate study, and members of the various demographic groups tend to favor some fields over others. The interactions of all these elements are ultimately reflected in measures of the elapsed time it takes to complete the doctorate.

Generally, time-to-doctorate, as measured by both TTD and RTD, has been considerably shorter for men than for women, shorter for temporary residents than for permanent residents and U.S. citizens, and among U.S. citizens, shorter for Asians/Pacific Islanders than for other racial/ethnic groups.

The gaps between women and men in TTD and RTD have narrowed over the years (figure 4-16 Figure.). Men's TTD and RTD increased 2 years between the periods 1960–64 and 1995–99, to 10 years for TTD and to 7 years for RTD, whereas women's TTD remained about the same and their RTD increased from 6 to 8 years. The differences in time-to-doctorate for men and women correspond to the differences in the median times for S&E and non-S&E fields. Compared with men, women were more concentrated in non-S&E fields than in S&E fields. Consequently, women's time-to-doctorate measures are more similar than men's to the pattern for non-S&E fields.

TTD and RTD increased for each citizenship group between 1960–64 and 1995–99, with the increases largest for permanent residents (figure 4-17 Figure.). Temporary residents, with their heavy concentration in S&E fields, took less time in 1995–99 to complete their doctorates than permanent residents and U.S. citizens did. An analysis of time-to-degree for various demographic groups, which incorporated information for students who did not complete degrees, showed that foreign Ph.D. students have higher completion rates as well as lower average times to degree completion (Espenshade and Rodriguez 1997).

Reliable race/ethnicity data first became available in 1975. Since then, among U.S. citizens the median RTD has increased for every racial/ethnic group, and the median TTD has increased for every group but Asians/Pacific Islanders (figure 4-18 Figure.). Among racial/ethnic groups in 1995–99, Asians/Pacific Islanders—highly concentrated in S&E fields—had the shortest median TTD (9 years) and the shortest median RTD (7 years). Blacks, with their heavy concentration in the field of education, had the longest times to degree completion.


Espenshade, T.J., and G. Rodriguez. 1997. Completing the Ph.D.: Comparative performances of U.S. and foreign students. University of Texas/Social Science Quarterly 72(2):593–605.

National Academy of Sciences, National Research Council (NAS/NRC). 1968–98. Summary Report 1967…1996: Doctorate Recipients from United States Universities. Annual series of survey results from the Survey of Earned Doctorates, conducted by NAS–NRC under sponsorship of the National Science Foundation, National Institutes of Health, National Endowment for the Humanities, U.S. Department of Education, and U.S. Department of Agriculture. Washington, DC.

National Opinion Research Center (NORC). 1998–2002. Doctorate Recipients from United States Universities: Summary Report 1997…2001. Annual series of survey results from the Survey of Earned Doctorates, conducted by NORC for the National Science Foundation, National Institutes of Health, National Endowment for the Humanities, U.S. Department of Education, U.S. Department of Agriculture, and National Aeronautics and Space Administration. Chicago.

National Science Foundation (NSF). 2000. Modes of Financial Support in the Graduate Education of Science and Engineering Doctorate Recipients. NSF 00-319. Arlington, VA.

U.S. Department of Education, National Center for Education Statistics (USED/NCES). 2002. Digest of Education Statistics, 2001, table 182. Washington, DC.


1   Some Ph.D.s in dual-degree programs may have received the professional degree after the doctorate, and other Ph.D.s may have later independently pursued a professional degree. Only professional degrees that were received by the time the doctorate was conferred are recorded in the Doctorate Records File, and thus only those professional degrees are included in the data presented in this report.

2   Information on all modes of support can be found in the annual summary reports of SED survey results (NORC 1998–2002, NAS/NRC 1968–98).

3   Before 1969, the calculation of RTD and TTD was based on whole years; from 1969 on it was based on months and years. The calculation of PTD is based on whole years.

Front Matter

Chapter 1. Introduction

Chapter 2. Historical Background

Chapter 3. Field Choices and Demographic Characteristics of Doctorate Recipients

Chapter 4. Path to the Doctorate

Chapter 5. Baccalaureate and Doctoral Institutions

Chapter 6. After the Doctorate

Appendix A. Detailed Doctoral Fields and Demographic Characteristics of Ph.D.s

Appendix B. Top 50 Baccalaureate and Doctoral Institutions and Doctorate Production by State

Appendix C. Technical Notes


List of Tables

List of Figures

Figure Reference Tables (web only)

Supplemental Tables (web only)


Susan  T. Hill,
Project Officer
Human Resources Statistics Program
(703) 292-7790
Previous ChapterPrevious Back to Top of pageTop NextNext Chapter

U.S. Doctorates in the 20th Century | NSF 06-319 | October 2006

National Science Foundation Division of Science Resources Statistics (SRS)
The National Science Foundation, 4201 Wilson Boulevard, Arlington, Virginia 22230, USA
Tel: (703) 292-8780, FIRS: (800) 877-8339 | TDD: (800) 281-8749
Last Updated:
Mar 18, 2011
Text Only