Our PhD Employment Problem, Part I

Our PhD employment problem is very simply described: there’s a mismatch between the number of graduate students earning doctorates each year and the number of tenure-track faculty positions available to them. There are too few tenure-track jobs for the PhD recipients who are qualified to compete for them. The problem is most commonly attributed to an overproduction of PhDs. But it can be understood and has been argued to be the result of an underproduction of tenure-track positions symptomatic of institutions’ increasing reliance on non-tenure-track faculty appointments.

Figure 1, with its red trend line tracking new PhD recipients and its blue trend line tracking jobs advertised in the MLA Job Information List, has formed our basic image of the depressed academic job market that has persisted since the 1970s. (Click on the chart to see a full-screen version. The curve representing PhD recipients is based on a query for graduates in “English and literature” from the United States Survey of Earned Doctorates [SED] data as maintained on WebCASPAR, a National Science Foundation [NSF] Web site of United States government data resources on higher education. “English and literature” is an NSF aggregate of several SED subfields, including English and American language and literature, creative writing, comparative literature, classics, folklore, speech and rhetorical studies, and, beginning in 2012, rhetoric and composition. The SED is the federal government’s annual census of new PhD recipients.)

Fig. 1
01_PhDs_Jobs

More than an image, this charting has served as our mental model for the job market and its troubles—especially the two episodes of acute crisis that become vividly apparent where the blue trend line for jobs plunges below the red trend line for PhD job seekers in the mid-1990s and most recently since 2008. I want to argue that this model is oversimplified and has led us to think too narrowly about PhD career paths, actual and possible.

Consider, for example, the record of initial employment placements across the fourteen studies of PhD placement that the MLA has conducted since 1977, as shown in figure 2. (The most recent study sought information about placement, as of November 2010, of graduates who received their degrees between 1 September 2009 and 31 August 2010.) This charting of MLA placement survey findings reminds us that the PhD employment (and unemployment) picture has always included several strands in addition to postsecondary faculty positions on and off the tenure track—including administrative posts in postsecondary institutions and employment outside higher education as well as postdoctoral fellowships.

Fig. 2
02_MLA_placement

The crisis episodes of the mid-1990s and since 2008 are clearly reflected in the sharp downturns in the percentage of graduates finding tenure-track placements and corresponding upturns in the percentage taking full- or part-time non-tenure-track positions. At the best, across the record of fourteen MLA surveys over three decades, just above half of a given year’s graduates have found placements to tenure-track faculty appointments in the year they received their degrees. In trough years, the figure has dropped below 40%. (Shown here are findings for graduates of doctoral programs in English. The picture for languages other than English is similar.)

Also striking in the record of the MLA placement studies is how unchanging and predominant placement to postsecondary teaching has been. Across the ups and downs of three decades, close to 80% of program graduates have consistently gone on to postsecondary teaching right after graduate school, whether as tenure-track assistant professors or as non-tenure-track instructors. The notable development has been an exchange of placements beyond postsecondary teaching in favor of placement to postdoctoral fellowships. Employment outside higher education, whether in K–12 education, government, not-for-profit organizations, or for-profit enterprises, has actually seen significant decline since the 1980s. The unemployment rate (graduates reported as not employed and seeking employment), which, after reaching 10.5% in the crisis of the 1990s dropped to 2.1% in 2006–07, increased to 4.2% in 2009–10. Again, what we are looking at here is placement of graduates directly after completing the PhD, within the same year a graduate received the degree. Interestingly, and in contrast to postdoc placements, placements to academic administration, where we might expect to see some indication of a turn to nonfaculty (or “alt-ac”) positions in higher education, has yet to break out on the MLA survey as a significant percentage of initial placements, reaching a peak at 3.9% in 2003–04 and dropping back to 2.0% in 2009–10.

The concentration of humanities’ graduates in postsecondary teaching positions immediately after completing doctoral study is striking in comparison with graduates in social sciences and sciences, as data from the SED make apparent (fig. 3). If a humanities degree recipient has a definite commitment to employment or postdoctoral study at the time they complete the SED questionnaire, chances are 8 in 10 that the commitment is to a postsecondary teaching job (whether on or off the tenure track), by far the highest among the five major disciplinary areas.

Fig. 3
03_SED_academia

Correspondingly, the SED data show how dominant postdoctoral study is in the sciences and also how the upward trend in placement to postdoctoral study that we saw in the MLA placement findings is shared across the disciplines (fig. 4). (Do the sharp increase in placements to postdoctoral study in science and engineering after 2008 and the equally sharp drop since 2010 reflect both early adjustments to the financial crisis on the part of programs and graduates and subsequent reduced allocations of government funding for research labs in these fields?)

Fig. 4
04_SED_postdoc

But how does the humanities’ overall placement record compare with that for the sciences and social sciences? Figure 5, drawn from the SED, compares the percentage of graduates in the different disciplines with definite commitments, whether to employment or postdoctoral study, across two decades, from 1991 to 2012. The SED consistently finds that the humanities have the lowest percentage of graduates reporting a definite postgraduation commitment to employment or postdoctoral study. But, no surprise, things have gotten a lot tougher for everyone since 2008. What may come as a surprise is the sharp declining trend for graduates in the life sciences and a trend for the humanities that is actually slightly positive (if by a considerable margin still lowest among the disciplines in the percentage of graduates reporting definite postgraduation commitments, although the distance between the humanities and life sciences has narrowed).

Fig. 5
05_SED_placement

It is illuminating to consider the numbers behind these percentage values (fig. 6). The life sciences have experienced dramatic increases in degree recipients, which have doubled in number since 1982 and show no sign of slowing down. Humanities degrees increased through the 1990s but, unlike the sciences, leveled off after 1997. (The most recent year for which SED data are available, 2012, did see a notable uptick in the number of humanities degree recipients.)

Fig. 6
06_SED_production

So the relationship between trends in numbers of graduates and trends in percentage of graduates finding postgraduation placements differs for the sciences on the one hand and the humanities and also education on the other. Figure 7 illustrates the story for life sciences over the two decades since 1992. (The blue and black trend lines for absolute numbers of graduates refer to the left axis; the red trend line for the percentage of graduates placed refers to the right axis.)

Fig. 7
07_SED_life sciences

The absolute number of life sciences graduates finding placements immediately after graduation has been increasing. But the number of graduates has been increasing a lot faster. So the percentage of graduates reporting definite postgraduation placements has seen a substantial decline of ten percentage points. It is now approaching 60%, where it had been 70% or above.

Compared with life sciences, PhD production in the humanities looks flat—as does the number of graduates finding placements immediately after graduation (fig. 8). And the two fields have actually gotten a lot closer in the percentage of graduates with definite commitments to employment or postdoctoral study than was the case ten or fifteen years ago.

Fig. 8
08_SED_humanities

Education shows yet another pattern (fig. 9), with a sharply declining number of graduates receiving degrees—down 25% since 2007—but accompanied by an equally abrupt decline in the number of graduates finding placements. Consequently, the substantial reduction in graduates with PhDs in education has not improved the percentage of graduates reporting definite commitments to employment or postdoctoral study (just under 70% in 2012).

Fig. 9
09_SED_education

But what about humanities PhD employment and careers in the long term? Figure 10 presents data from the 1995 Survey of Doctorate Recipients (SDR), a United States government survey and companion to the SED that tracks a sample of PhDs from the time they receive their degrees to age seventy-five. In 1995, the last year data were collected for the humanities, the SDR found something on the order of 40% of the entire humanities PhD population working in positions other than postsecondary teaching.

Fig. 10
10_SDR_outcomes

Humanities participation in the SDR, which was funded with support from the National Endowment for the Humanities, was lost when the endowment’s budget was cut in the mid-1990s (data are still collected for the sciences). The lack of current information about the employment sectors where humanities PhDs find work and what they are actually doing in their work has prompted several recent efforts, including projects being undertaken by the MLA and the American Historical Association (AHA) with support from the Andrew W. Mellon Foundation to track the current occupations of post-1995 PhDs and the 2012 survey Katina Rogers directed for the Scholarly Communication Institute (SCI). The AHA has issued a report of findings from its study of PhD employment outcomes. The (very preliminary and subject to change) returns from the MLA’s effort are so far very much in line with the AHA’s findings for history PhDs, as shown in figure 11. A little under 60% of the PhDs we’ve located are working as tenured postsecondary faculty members. Just over 10% are in non-tenure-track faculty positions. Between 6% and 7% hold professional staff positions in postsecondary institutions; between 4% and 5% are deans, provosts, or college presidents (the “senior postsecondary administration” category on this chart); and a not insignificant 21% are working outside postsecondary education altogether. (These early results from the MLA’s research come from a subset of several hundred PhDs, all of whom earned degrees before 2000; consequently, for the purposes of this first, quick take, we did not attempt to distinguish tenure-track assistant professors.)

Fig. 11
11_MLA_prelim

It’s interesting to look at these findings in relation to findings from the SCI’s 2012 survey, shown in figure 12, since that effort focused specifically on people with graduate degrees who were working in “alt-ac” positions—pursuing careers in postsecondary institutions but not as faculty members. (The SCI findings include ABDs and respondents who left graduate school with master’s degrees as well as those who completed the PhD.) Asked about the career they expected to enter at the start of graduate school, the SCI respondents overwhelmingly indicated a career as a tenured faculty member.

Fig. 12
12_SCI-1

And an equally overwhelming percentage of the SCI respondents said that when they started graduate school they were either certain or fairly certain that they would pursue a career as a tenured faculty member (fig. 13). Remember, for most of the survey respondents, their actual careers have taken them in other directions.

Fig. 13
13_SCI-2

So the evidence we possess points to two conclusions: people who enter the long and arduous path of doctoral study in the humanities do so having a postsecondary faculty career as their primary goal, and people who pursue graduate education in the humanities actually find careers in a far broader range of professional positions than postsecondary teaching, even if their first job after graduate school is a postsecondary faculty position on or off the tenure track. So the question isn’t whether doctoral study can lead to careers beyond postsecondary teaching—it already does and has for decades. The question comes down to the view doctoral programs and their inhabitants take of what has long been the simple fact.

But, given that graduate students in the humanities declare becoming a postsecondary faculty member as their predominant career goal and motive for undertaking doctoral study, it seems pertinent to our discussion to remind ourselves what is happening to the faculty. That is the topic for The Trend‘s next post.

David Laurence

More on That Northwestern Study: The Authors Speak, but Is Anyone Listening?

“Are Tenure Track Professors Better Teachers?” The three authors of the NBER working paper—David Figlio, Morton Schapiro, and Kevin Soter—spoke on 7 October at a colloquium organized by Northwestern University’s Institute for Policy Research (IPR), which Figlio directs. (A PDF of the paper is available for free at the IPR Web site.) As reported in the 8 October issue of the Daily Northwestern, their remarks underscore the point made in my blog post on their paper.

From Schapiro, president of Northwestern University: “In retrospect, I wish we had been a little clearer about the fact that we have non-lined faculty, but most of them are regularly renewed longtime professors here teaching full time for us and have been doing it forever. It’s a big difference [from] when I was at (the University of Southern California) and we called them ‘freeway flyers.’”

From Figlio: “The biggest takeaway message, in my opinion, is really [that] major research universities have these outstanding teachers teaching classes and we need to give these outstanding teachers more respect. We in the professoriate and maybe NU as an institution, but not just NU, need to recognize that if these people are doing such a good job in the classroom, maybe they should be even more integrated into the fabric of great research universities.”

New York TimesWall Street JournalChronicle of Higher EducationInside Higher Ed, and university and college presidents from sea to shining sea, are you listening?

David Laurence

One-Tenth of a Grade Point

The trend toward a majority non-tenure-track faculty has begun to attract serious economic analysis as—possibly—an institutional problem. You know something is up when administrators seek evidence that no damage ensues to students’ education from staffing practices and personnel policies that have created a higher education faculty with, in four-year institutions, only a third of its members employed in full-time tenured or tenure-track faculty positions (and under 15% in two-year colleges).

Enter the working paper recently released by the National Bureau of Economic Research, “Are Tenure Track Professors Better Teachers?,” by David N. Figlio, Orrington Lunt Professor of Education and Social Policy and of Economics at Northwestern University; Northwestern University’s president, Morton O. Schapiro; and Kevin B. Soter, from the Chicago-based consulting firm The Greatest Good. The paper examines first-term classes taken by first-year Northwestern University students across eight student cohorts, from fall 2001 to fall 2008, some taught by faculty members inside the tenure system and some taught by faculty members outside the tenure system. The authors ask two questions: Did the likelihood that a student took a subsequent class in the same subject vary by the tenure status of the instructor in the first course? Did the grade the student received in the subsequent course vary by the tenure status of the instructor in the first course? The paper states their core findings as follows:

[A] non-tenure track faculty member increases the likelihood that a student will take another class in the subject by 7.3 percentage points (9.3 percentage points when limited to classes outside the student’s intended major) and increases the grade earned in that subsequent class by slightly more than one-tenth of a grade point (with a somewhat greater impact for classes outside of the intended major). (9)

The conclusion that follows most directly from these findings should come as welcome news to MLA members, especially in the light of the exceptional character of the study’s institutional setting and the specific courses on which it focuses. It should surprise no one to have evidence that a cadre of full-time non-tenure-track faculty members, most of whom have what the paper’s authors describe as “a longer-term relationship with the university” (9n8), perform well as teachers in the introductory courses that are central to the work Northwestern University hires them to do. Or that these full-time non-tenure-track instructors even improve—slightly—on the support colleagues inside the tenure system can offer students in these introductory courses, courses that are not as central to the work, including the teaching work, Northwestern University hires faculty members inside the tenure system to do. Perhaps the surprise should be how small the differential is rather than how large. We’re talking about one-tenth of a grade point on a four-point scale.

Of course, the study and its conclusions will be put to polemical uses that can be anticipated to respect none of the features that ought to make the study welcome and to indulge all the phony generalizations that are bound to make its findings damaging. Under cover of the conveniently elastic term “adjuncts,” results of a study confined exclusively to full-time faculty members are sure to be cited as showing what the findings emphatically do not show: that the widespread institutional abuse of part-time faculty members does no damage, whether to students’ education, to the faculty, or to institutions. Memo to the media: the findings of this study have nothing to do with faculty members whose status or function can properly be described as adjunct, even if institutions use a title like “full-time adjunct professor” to categorize them.

The authors are curious to understand how their results may speak to the debate about higher education’s staffing practices and the dramatic expansion of the segment of the faculty employed off the tenure track. They note that in the United States in 1975, “57% of all faculty [members] (excluding graduate students) were in the tenure system; by 2009 that figure had been cut almost in half to 30%” (2). Interestingly, according to the United States Department of Education’s Employees by Assigned Position Survey (EAP), in fall 2011 tenured and tenure-track faculty members made up 57.7% of all non-medical-school faculty members at Northwestern University, compared with 28.2% across all degree-granting institutions, both two- and four-year, and 33.9% across the four-year institutions. With respect to tenured and tenure-track faculty members, the faculty demography of Northwestern University today is comparable to the faculty demography prevailing across all institutions of higher education in the United States in 1975.

Moreover, according to the EAP, 64.8% of the Northwestern faculty outside the tenure system was full-time in fall 2011.  Among four-year colleges and universities, 70.3% of the non-medical-school faculty outside the tenure system was part-time in fall 2011, nearly the reverse of conditions at Northwestern. In the light of the decidedly exceptional demography of the Northwestern University faculty, efforts to generalize the results of this study to the wider universe of four-year colleges and universities are untenable or at best premature. Unless, of course, a right understanding of this study and its institutional setting really argues for returning the faculty to the conditions of tenure status and full- and part-time employment status prevailing across the system in 1975 and still prevailing at Northwestern today.

The most consequential implication of the paper’s analysis may be stated as follows. If non-tenure-track teachers are professionals who perform well in the work they are hired to do, shouldn’t this—now majority—faculty be included in institutional employment and compensation policies that treat them as professionals? Instead, it seems depressingly clear that the paper will be abused to rationalize the further immiseration of the large and rapidly expanding part-time segment of the non-tenure-track academic workforce, under the excuse that “the research shows” faculty members outside the tenure system teach at least as well and often better, irrespective of the size of their classes, their course loads, their compensation, their full- or part-time employment status, the working conditions they endure, or how little support they receive. Which leads to the question, How will the authors, most especially the president of Northwestern University, respond if their findings are used to assert such unjustified conclusions?

David Laurence

Mismeasuring the Humanities

In the wake of the release of The Heart of the Matter, a report commissioned by the American Academy of Arts and Sciences that argues for the importance of the humanities and social sciences, and a trio of Harvard reports advocating humanities study, sympathetic voices have come forward to lament a purported decline in humanities study, to assess the cause of such a decline, and to prescribe remedies. Yet, as Ben Schmidt has pointed out, conclusions about trends in humanities degrees need to take into account broader changes in higher education. Citing the percentage of undergraduates receiving degrees in the humanities in, say, 1971, compared with the percentage receiving humanities degrees today omits at least two key parts of the picture.

1. The percentage value for “then” is inflated by the high percentage of women who were, until the 1970s, a semicaptive audience for study in the humanities (and also in education).

For example, in the late 1960s, nearly 12 of every 100 degrees earned by women were in English. For men, the figure was about 4.5 of every 100 degrees. By 1983, just over 4 of every 100 degrees women earned were in English, and the gap in the number of English degrees awarded to women versus men shrank from over 7 of every 100 degrees to less than 2. That is, as women gained entry to professions formerly closed to them, their choice of English as a major came to resemble men’s more closely, and the proportion of degrees they earned in English dwindled.

The other side of the coin appears in business degrees. In 1966, fewer than 3 of every 100 degrees earned by women were in business, compared with 20 of every 100 degrees for men. Twenty years later, in 1986, the figure reached 22 of every 100 degrees for women, and the gap between the rate at which women and men chose business as a field of study narrowed by 12 percentage points (over 70%), from 17 to 5 percentage points.

2. The percentage value for “now” can be artificially depressed if it is calculated using a too-restrictive aggregation for the humanities.

The 7% figure often cited as the percentage of college graduates who majored in the humanities in 2010 reflects a National Science Foundation aggregation that, for example, categorizes art history degrees with the arts rather than with the humanities and degrees in ethnic studies, women’s studies, and other area studies fields with the social sciences. While the assignments are debatable, a percentage value derived from a system that places them all outside the humanities should not be taken simply at face value.

Using a more capacious aggregation that includes academic studies in the arts and in humanities-connected programs in area studies, the Humanities Indicators has created an accounting showing that degrees in the humanities have steadily held a share between 10% and 12% for more than two decades.

A decline in the percentage of humanities degrees earned did occur between 1970 and 1986, but the decline affected all the liberal arts disciplines, not just the humanities, and it was short-lived. The humanities share increased from 1986 to 1990, and since 1990 it has remained unchanged for two decades. Taking the two dates 1966 and 2011 while completely ignoring the middle gives an inaccurate picture.

A numerical decline is not prima facie evidence sufficient to prove there has been a “fall” or “demise.”

A list of articles related to the recent reports appears on the From the President blog.

David Laurence