[PDF version]

As we look back over the past 40 years, we find that after decades of lowering the barriers of class and privilege, the ranks of new history PhDs are growing less diverse and more likely to draw from a narrow range of elite institutions.
A close analysis of information contained in the federal Survey of Earned Doctorates (SED) from 1966 to 2002, and a database of 11,562 history PhDs reported to the AHA Directory of History Departments, Historical Organizations, and Historiansfrom 1990 to 2004, suggests that

• The doors to graduate study in elite private universities are largely closed to students who received their degrees in public colleges and universities outside of a few with top-tier PhD programs.

• Receiving an undergraduate degree from an elite school appears to be an important marker of future success in the academic job market.

• A majority of history PhDs did their undergraduate studies in departments that also grant the PhD—with almost a quarter of the students entering from the top-tier programs (as ranked by the National Research Council in 1993).

• Beyond the PhD-granting history programs, a small number of private liberal arts colleges played a critical part in feeding undergraduates into the pipeline of future history PhDs.

• After differing significantly in their undergraduate origins during the late 1960s and early 1970s, minorities and women receiving history PhDs became increasingly similar educationally to their white and male counterparts.

• There is much less geographical mobility than one might expect. Almost half of the U.S. citizens receiving history PhDs over the past 15 years received their degree in the same region (a quarter from the same state) as their undergraduate degree. 1

History in Time and Academia

It helps to see these trends in a broader context—both temporally and in relation to other disciplines—because they demonstrate both a marked change over the past 15 years, and how history differs from other fields. Of particular note is the place of private colleges and universities as the source for future history PhDs (Figure 1).2 In 1966, 57 percent of the new history PhDs had an undergraduate degree from a private college or university, while just 37 percent entered from a public institution. The balance, barely 8 percent, received their degrees from foreign institutions. Since that time, the proportions have narrowed significantly—even reaching parity for a time in the late 1980s. Through the 1990s, however, the gap between public and private colleges and universities began to diverge again, growing to a 47 percent to 42 percent advantage for private institutions.

On all these measures, history is markedly different from the social sciences (where it is still often lumped in the federal data). Among social science fields, students with degrees from public institutions have comprised a large plurality of the new PhDs throughout the past 30 years, and a sizeable majority of the students had degrees from domestic institutions. Less than one-third of the social science PhDs in the latest cohort had their undergraduate origins in a private institution.

History is more closely aligned with the humanities in these trends, though history PhDs are still more likely to come from private institutions. Among all humanities PhDs, the proportion from private institutions fell from 56 percent in the first cohort, to 44 percent in the most recent cohort (lingering as low as 41 percent in the 1989 to 1994 cohorts).
The patterns are similar among students with undergraduate degrees from foreign colleges and universities. In the 1968–70 cohort, just 6 percent of the new history PhDs had baccalaureate degrees from foreign institutions. Degree recipients with foreign undergraduate degrees comprised over 13 percent of the new doctorates in the 1992–94 cohort, before slipping a bit to just over 10 percent among the latest cohort.3

Here again, history is much more closely aligned with the humanities and significantly different from other social science fields. Among the most recent cohort, 15 percent of the new humanities PhDs had degrees from foreign institutions, while 30 percent of social science PhDs had their undergraduate degrees from abroad.

A Core of Programs

Within these broad trends, one of the most notable changes among the domestic sources for history doctoral students is the resurgence of a limited number of elite programs (Figure 2).

Among history PhDs in the 1968–70 cohort, 25 schools supplied 25 percent of the students receiving history PhDs, a proportion that fell to 20.8 percent in the 1983–85 cohort. In the latest cohort, however, the "market share" for the top 25 feeder schools had risen back to 26.5 percent of the total. 4 The trend is similar if one looks at the top 100 U.S. institutions feeding undergraduate students into the history PhD pipeline. The leading 100 schools supplied almost 53.3 percent of the students in the late 1960s, and fell to below 49 percent through much of the 1980s, but rose steadily through the 1990s to almost 55 percent of the latest cohort.

It is important to note that the resurgence of the core programs occurred even as the total number of institutions sending students on to history PhD programs was growing. According to the federal data, the 1986–88 cohort came out of 530 programs, as compared to 633 programs for those receiving degrees in 2000–03. As a result, dispersion beyond the core programs has become much wider.

This dispersion is reflected in the disparity between the demographics of all students who receive history BAs and the undergraduate origins of recent history PhDs. The baccalaureate demographics of those receiving the history PhD were markedly different from the larger universe of students receiving bachelor's degrees in history. Among all students receiving bachelor's degrees in history in 2001, the top 25 programs accounted for 18.2 percent of the baccalaureate degrees in the discipline, while the top 100 programs produced 41.2 percent of the history bachelor's degrees—more than a third lower than their representation among history PhDs. Even as the core programs were gaining in market share among new history PhDs over the past 15 years, the dominance of those schools in the overall production of history BAs was diminishing.

Factors for Change

There are a number of possible reasons for these recent trends. The expanding pool of undergraduate programs represented among new history PhDs is at least partially a function of the growing number of programs conferring the PhD throughout the 1990s—there are 154 programs in history today, as compared to 136 in 1990 (and 105 in 1970). Given the tendency of these programs to draw students from schools in their local area (detailed further below), the increased geographic spread of the new programs served to expand the entry points for a small number of undergraduate students.

Less tangibly, these numbers only tell us about who completes the PhD—not who is actually admitted to a program. This leaves open the question of whether undergraduate study at one type of institution provides a better preparation for the rigors of history doctoral study. To the extent that the 30-year trends indicate a shift first toward and then away from greater diversity in baccalaureate origins, however, this seems like a less compelling issue—unless one assumes that undergraduate training at public and non-elite institutions has become less capable of preparing students for history studies over the past 20 years.

The more likely force for concentration is the growing effort—particularly in the top-tier private institutions—to increase the funding commitments for their graduate students. The AHA's Committee on Graduate Education (CGE) speculated that this change would lead to increased caution among those admitting students into PhD programs conferring large aid packages, and a tendency to use the undergraduate institution of origin as a key criterion in selection. The CGE feared this move would lead to a decrease in the institutional diversity of the student population. 5

A closer look at information on new history PhDs submitted to the AHA's Directory seems to validate those concerns.6 Among historians in the 1990–94 cohort those who received their PhDs at private universities in the top-quarter of the NRC rankings, the 10 top feeder schools supplied 31.2 percent of the new PhDs. Among PhDs conferred at those same schools in the past five years (2000–04), the top 10 feeder schools accounted for 36.6 percent of their baccalaureate degrees. In comparison, among public institutions in the top tier of the NRC rankings—which generally lack comparably large funding packages—the concentration of students from the top ten feeder schools had barely changed, growing from 23.1 percent to 23.7 percent between the two cohorts.

This increasing concentration was also evident in the changing number of undergraduate programs represented among the new PhDs. At the top-tier private programs, the number of undergraduate institutions represented fell 17.9 percent (from 234 in the 1990–94 cohort to 192 in the 2000–04 cohort), even as the number of PhDs conferred by those programs rose 12.9 percent (from 673 to 760 between the two cohorts). In contrast, as the number of PhDs conferred by public programs in the top tier rose 34 percent (from 614 to 823 between the two cohorts) the number of undergraduate institutions increased by 4 percent (from 270 to 280).

The Institutional and Geographic Mobility of History PhDs

To help demonstrate some of the particular characteristics of the core programs feeding students into history PhD programs, we pulled out the top 25 programs feeding students using two different measures.

The first measure, using the gross numbers of undergraduate students from the institution receiving PhDs, produces a list we label the "Big 25" (Table 1). These institutions provided the baccalaureate degrees for 24.8 percent of the history PhDs conferred between 1989 and 2003. This list is dominated by large state universities on the two coasts, together with a few of the top-ranked history PhD programs at private universities. Institutionally, 23 of the programs are at major research universities, with only two liberal arts colleges on the list.

To adjust for programs that show up on the former list based on the sheer size of their programs, we constructed a second list of programs, the "Select 25" (Table 2), which conferred the highest proportion of baccalaureate degrees among new history PhDs relative to the actual number of history bachelor's degrees they had conferred.7 Schools on this list provided the baccalaureate degrees for 15.4 percent of the history PhDs conferred between 1989 and 2003.

This list proved to be quite different, with only nine programs showing up on both lists. The "Select 25" list is dominated by private liberal arts colleges (16 in all), with much of the overlap comprised of top-ranked history PhD programs at private universities.

Using the Directory data, the significance of the key producers on the Big 25 and Select 25 lists becomes more apparent because it allows us to connect the PhD recipients' undergraduate school to the university conferring the PhD, and also allows us to test whether this made any difference in placements on the academic job market.

Not surprisingly, the top tier of PhD schools in the latest National Research Council rankings draw heavily from both the Big 25 and Select 25 schools. Programs ranked in the top quarter of the rankings drew an average of 23.5 percent of their PhDs from the Select 25 schools, and 33.4 percent of their PhDs from the Big 25 schools. In comparison, among PhD recipients from all other institutions, only 5.9 percent came from the Select 25, and 10.0 percent came from the Big 25 institutions.

As we have noted in past reports, there is also a marked difference in the gross number of students receiving degrees from the top-tier schools, which confer almost half of the history doctorates annually. So, for instance, the top 26 institutions in the NRC rankings graduated 1,310 PhDs with undergraduate degrees from the Select 25 schools over the past 15 years, as compared to 353 from among the other 122 institutions combined.

This disparity points to two significant differences between the top-tier schools and the other departments conferring PhDs—their tendency to draw from among their own ranks, and their national (and indeed international) reach.
The AHA's data indicates that programs in the top tier of the NRC rankings had the greatest concentration of students from a limited number of elite programs. While numerically a larger number of the undergraduates from these institutions come from programs further down the rankings, this was a function of the much larger numbers of students receiving PhDs.

The top–ranked programs drew students from 604 different undergraduate programs in the United States, while programs in the bottom half drew students from 520 institutions. But again, with nearly twice the number of students receiving degrees from the top-ranked universities, the distribution of undergraduate origins was much more even at the bottom-ranked institutions. Top-tier programs privileged students from other elite institutions, drawing more than a quarter of their PhDs from a top-tier university—either the same university or one of the other 26 programs at the top of the rankings. In contrast, only 11.3 percent of the students receiving degrees from programs in the second tier had a baccalaureate degree from a top-tier program, and that proportion falls to just 7.6 percent among PhDs from other universities.

One of the other important barriers seems to be the divide between public and private institutions. If we look just at students with undergraduate degrees from colleges and universities in the U.S., private institutions, particularly among the top-ranked programs, appear much less likely to draw in students from public colleges and universities. Among PhDs from the top-tier private institutions, 71.9 percent of their PhDs had undergraduate degrees from other private institutions (Figure 3). In comparison, at public universities in the top tier, barely half (52.9 percent) came from private institutions. The differences between public and private institutions becomes slightly less pronounced the further down the rankings one looks, but a clear divide between public and private institutions remains evident.

Alongside the apparently limited institutional mobility (and modestly contributing to it), one of the more intriguing findings of the survey is the relatively limited amount of geographic mobility between the school that conferred the undergraduate degree and the university conferring the PhD. Among those with an undergraduate degree from an institution in the U.S., a bit more than 41 percent received their PhDs in the same region as their baccalaureate degrees—and more than a quarter (28.1 percent) received their baccalaureate degrees in the same state as their PhDs.
The apparent anomalies in this dataset are from a few states—such as Rhode Island, New Jersey, and the District of Columbia—that have a number of top-tier research universities across state borders but within 50 miles of their largest degree-granting institutions. In comparison, among PhDs in both California and Hawaii, more than half (54 percent and 61 percent, respectively) of the new PhDs were receiving their degree in the same state.

However, there was a marked contrast in the national reach of top-tier PhD-granting institutions. Less than a third of the PhD recipients at the top-tier programs were receiving their degree in the same region, as compared to more than half of the new PhDs in the bottom half of the NRC-ranked programs.

As one might expect, internal recruitment is an important factor in this. Among all history PhD recipients 8.4 percent received their degrees from the same institution that conferred their baccalaureate degrees. There was very little variation in this. Programs in the top tier were just a bit less likely to confer the PhD on someone with a bachelor's degree from the same institution, as 7.3 percent of the PhD recipients in those programs had bachelor's degrees from the same university. The only marked variation in this trend was among programs that did not exist—or were not ranked—when the NRC did its study in 1993. In these new and unranked programs, 12.4 percent of the PhDs were receiving the degree from the same institution.

In terms of international reach, the survey found less variation among the programs. Departments in the second tier of the NRC rankings graduated a slightly higher proportion of PhDs with baccalaureate degrees from foreign colleges and universities—13.1 percent, as compared to 12.4 percent at programs in the top tier. At programs in the bottom half of the rankings, however, just 8.4 percent of the PhDs had undergraduate degrees from foreign institutions.

Undergraduate Origins of Women and Minority History PhDs

The apparent continuity in the broad trends masks significant changes in the underlying demographics of history PhDs, particularly in terms of the undergraduate institutions that have served as feeder schools for women and minorities.

It is well known that women and minorities receive history PhDs in much smaller proportions than their numbers in the general population, or even among students receiving bachelor's degrees in history.8 Despite numerous studies about why women and minorities are entering the "pipeline" of history PhDs in smaller numbers, the reasons for this difference remains unclear. What this survey does show, however, is that despite continuing differences in representation, the undergraduate origins of women and minorities have become increasingly similar to their white and male counterparts.

One way of measuring this is to look at the proportions of women and minorities entering PhD programs with baccalaureate degrees from the Select 25 and Big 25 schools (Figure 4). In the late 1960s, the Select 25 schools were a disproportionate source of women in history PhD programs, reflecting the importance of private liberal arts colleges (and particularly the "Seven Sisters" colleges—five of which are in the Select 25) in the education of women. In the 1968–70 cohort, 19.6 percent of new female PhDs received their undergraduate degrees from Select 25 schools and 18.7 percent had received their undergraduate degrees from a Big 25 school. In comparison, among their male counterparts only 12.2 percent had received their degrees from the Select 25 schools, while 20.3 percent had taken undergraduate degrees in the Big 25.

Since that time, however, the Select 25 programs have diminished as a source of women history PhDs, and in the early 1980s the Big 25 schools surpassed the Select 25, conferring 25.3 percent of the baccalaureate degrees of female history PhDs in the latest cohort, while the Select 25 conferred just 17.8 percent. In contrast, among male history PhDs, the proportions were nearly unchanged from 30 years before, as the Big 25 schools supplied 22.6 percent of the latest cohort and the Select 25 supplied 11.9 percent.

The federal government only began counting degree recipients by race in 1973. Looking at that cohort, we see a sharp disparity between the undergraduate origins of minority citizens and all history PhDs. In the 1974–76 cohort, only 7.4 percent of the minority history PhDs had received undergraduate degrees at Select 25 schools, while 14.9 percent had degrees from the Big 25 schools.

This has changed profoundly over the past 30 years. In the early 1980s, the proportion of minority students from the Big 25 programs fell sharply, before slowly climbing back up to surpass the average for all history PhDs. In the most recent cohort, 27.0 percent of minority PhDs received their undergraduate degrees from Big 25 institutions, and 16.4 percent had baccalaureate degrees from the Select 25 schools—slightly above the proportions among all history PhDs.
These mark much broader trends in the undergraduate origins of women and minority history PhDs. Since the early 1970s, research universities have become a dominant source of undergraduates receiving history PhDs in both categories. Among women, they have grown from 37.7 percent to 43.8 percent of the new PhDs, while baccalaureate institutions have declined as feeder schools for women, from 31.1 percent to 20.6 percent (slightly below the average for all PhD recipients, which stands at 22.0 percent).

Among new minority PhDs, research university BAs have grown from 35.5 percent in the 1974–76 cohort to 50.0 percent in the most recent cohort. Most of the shift was from master's level colleges and universities, which fell from 34.7 percent to 14.1 percent over the same period.

One of the other significant trends over the period is the shift from historically black colleges and universities (HBCUs) as the undergraduate origins of new African American history PhDs. The HBCUs conferred 42.9 percent of the baccalaureate degrees for African American in the 1974–76 cohort, as compared to 22.5 percent of the degrees for the most recent cohort.

With that as an exception, it appears that the differences between the undergraduate programs feeding women and minorities into PhD programs have all but disappeared.

Does It Make a Difference on the Job Market?

Obviously a wide range of factors play into a candidate's success on the job market, but using the limited comparison to the faculty listings in the Directory suggests that the undergraduate origins of a student could serve as an important marker of future success.

As we noted in our review of how many of the 1990–2004 PhDs had found jobs in the departments and organizations listed in the Directory, programs in the top tier of the NRC rankings appear more successful in placing their students—43 percent in comparison to almost 28 percent at all other programs.9

When we factor in the undergraduate origins of these PhDs, graduates from the Select 25 schools had a placement rate almost 27 percent higher than students from other programs (Figure 5). The institution that conferred the PhD still made a critical difference, but the undergraduate differential was evident among graduates from every rank of PhD-granting institutions. Almost 52 percent of the students with PhDs from the top-tier schools who also had undergraduate degrees from the Select 25 found full-time employment in one of the Directory-listed institutions. This compares to 44 percent among PhDs from the top-tier schools without Select 25 BAs. At all other programs, 33.1 percent of new history PhDs with undergraduate degrees from elite schools found employment in the Directory-listed institutions, as compared to 24.8 percent among those with bachelor's degrees from all other institutions.

Among PhDs with a degree from one of the Big 25 schools, the differential was perceptible, but not quite as pronounced.

Among PhDs from the top-tier institutions, 44.8 percent of those with Big 25 bachelor's degrees found full-time employment in Directory-listed institutions, as compared to 33.7 percent of those without such a degree. Among the other PhD programs, 30.8 percent of those with Big 25 degrees were placed, as compared to 24.6 percent among those without.

This is not to argue that the job market is closed to students from other programs—given the high level of competition in the history job market, the differences could easily be more pronounced if the ranks were truly closed. It does, however, serve as an important reminder that the beginnings of an academic career can play an important role in the way it ends.

— Robert B. Townsend is AHA assistant director for research and publications and a PhD candidate at George Mason University. Richard Bond and Liz Townsend assisted in the collection of dissertation data from the Directory. Thomas Bender, Chris Golde, Anthony Grafton, Roy Rosenzweig, and Peter Stearns all offered helpful readings of earlier drafts of this report.

Notes

1. Except where information is noted as having been drawn from data in the AHA Directory, all tabulations come from the National Science Foundation's WebCaspar system at http://caspar.nsf.gov/. The WebCaspar system provides tabulations of data at the department and subject level across a number of key demographics from both the SED (for doctorates) and the Department of Education's Integrated Postsecondary Education Data System (IPEDS) for information on baccalaureate degrees. For the analysis in this article, data tables from the WebCaspar system were merged with demographic data from the Department of Education's National Center of Education Statistics IPEDS Dataset Cutting Tool (hd2002) available at http://nces.ed.gov/ipedspas/dct/index.asp. The categories and classifications (of the educational institutions) used in this article are derived from the 2000 Carnegie Foundation for the Advancement of Teaching classifications available at http://www.carnegiefoundation.org/Classification/downloads.htm and the National Research Council's rankings of history doctorate programs published in Research-Doctorate Programs in the United States: Continuity and Change (1993) available at http://books.nap.edu/html/researchdoc/appendix_m.html.
The SED data allows us to tabulate the demographic profiles of students as they emerged by year, while the AHA data allows us to connect information about the undergraduate institutions to the type of institution conferring the PhD.
This article draws on methods and questions from earlier reports focused primarily on science and engineering doctorates, particularly Jeffrey A. Groen and Michael J. Rizzo, "The Changing Composition of American Citizen PhDs," paper presented at Science and the University conference, Cornell Higher Education Research Institute, Ithaca, NY, May 20–21, 2003, online at http://www.ilr.cornell.edu/cheri/; Ana Maria Turner Lomperis, "The Demographic Transformation of American Doctoral Education," Research in Labor Economics 13 (1992), 131–213; Philip J. Cook and Robert H. Frank, "The Growing Concentration of Top Students at Elite Schools," in Studies of Supply and Demand in Higher Education, ed. Charles T. Clotfelter and Michael Rothschild (Chicago: Univ. of Chicago Press, 1993), 121–40; and Carol H. Fuller, "Ph.D. Recipients: Where Did They Go to College," Change (November/December 1986), 42–51.

2. To develop time series data, we have cut the WebCaspar information into 12 cohorts of 5 years each. These serve to flatten out unusual annual fluctuations, but also usefully divide a number of key moments in the ebb and flow of history PhDs over the past 35 years. The first two cohorts (1968–70, 71–73) mark the explosion of PhDs that produced a surfeit of new PhDs in the early 70s. The next two cohorts (1974–76 and 77–79) marked a sharp reversal in the number of new history PhDs as programs cut back and students started to avoid the degree. The next three cohorts (1980–82, 1983–85, and 86-88) mark the resulting nadir in the number of new history PhDs. The next cohort (1989–91) finished during a transitional moment when the number of jobs (briefly) seemed to exceed supply and concerns about a "shortage of PhDs" in the humanities became current, and programs began to actively seek large numbers of applicants. The following two cohorts (1992–94 and 1995–97) completed their degrees as these concerns failed to bear out, and a new job crisis in the field took hold and programs again began curtailing the number of new students. Despite the cutbacks, the final two cohorts (1998–2000 and 2001–03) mark a peak in the number of new PhDs rein the recent period of growth in the number of new history PhDs resulting from sharp increases in admissions and the opening of new programs into the early 1990s.

3. Among the 11,563 history PhD recipients reported to the AHA Directory since 1990, 1,220 could be identified as having degrees from foreign schools. Not surprisingly, Canada was the largest source of undergraduate degrees (having conferred 201 of the baccalaureate degrees), followed by China (154), the United Kingdom (120), and Germany and Japan (with 50).

4. The top 25 (and top 100 schools below) were tabulated and assigned for each cohort. Only nine programs were among the top 25 programs in every cohort, and 67 programs made it into the top 25 for at least one cohort. In the tabulation of the top 100 programs, 28 institutions appeared on the list for every cohort, and 233 institutions appeared at least once.

5. Thomas Bender, et al., The Education of Historians for the Twenty-first Century (Urbana-Champaign: University of Illinois Press, 2004), 71–73.

6. Data gathered from information published in the AHA Directory of History Departments fall 1990 to fall 2004. Due to the slightly different time span available, we used five-year cohorts 1990–94 (N=3,055), 1995–99 (N=4,118), and 2000-04 (N=4,390) to provide time series of this data. Except where noted, these tabulations exclude foreign PhDs and those whose undergraduate degrees either were not listed or could not be identified (16.5 percent of the total).

7. The comparison was made using institutional-level data on history degrees completed from the IPEDs system for the last 14 available years (1987 to 2001, missing 1999) and the last 14 available years from the SED (1990 to 2003 inclusive) taken from the WebCaspar system. To correct for some smaller programs that could enter because of degrees from other programs, the list was further limited to programs that had conferred an average of at least 10 BAs over that span, and credited as the undergraduate degree for 15 history PhDs. While the University of Puerto-Rico, Rio Piedras could be included under these criteria, due to a sizeable difference in their numbers reported to the Directory, and their anomalous role on a range of other criteria, they are excluded from this list.

8. See most recently, Robert B. Townsend, "History Degrees Declining Relative to Other Fields, but Newer Data Provides Some Cheer," Perspectives(March 2005), 9.

9. Robert B. Townsend, "Job Market Report 2004," Perspectives 43:1 (January 2005): 13–19.