Opportunity Costs of the PhD: The Problem of Time to Degree

The road to a PhD in a humanities discipline is long—for graduates who completed degrees in 2012, 9.0 years long, the median time from first entry into graduate school to receipt of the degree (Doctorate Recipients, table 31). That nine-year median is not just long in itself; it is significantly longer than the median for 2012 graduates in social sciences (7.7 years), life sciences (6.9 years), physical sciences (6.7 years), and engineering (6.7 years). Only 2012 graduates in education, with a median 11.8 years to degree, take more time to complete their degree programs than graduates in the humanities.

As long as a nine-year path to the PhD may be, the class of 2012 humanities degree recipients took less time than any classes since the 1970s. The 2012 median of 9.0 years is notably shorter than the 9.5 years recorded by humanities graduates who received degrees over the five years 2006–10, and it is strikingly shorter than the record-high 10.7-year median for graduates who received humanities degrees between 1986 and 1990. (The median for 2011 graduates was 9.3 years, in case you were wondering.) In fact, since 1990, median time to degree (measured from year of entry into graduate school) has been falling across the disciplines. Figure 1 shows the history of time to degree in the four broad disciplinary areas of the arts and sciences over the fifty years from 1961 to 2010. The source is a custom report that NORC, the Survey of Earned Doctorates (SED) contractor, developed for the MLA from the raw SED data on time to degree from 1961 to 2010. The custom report grouped degree recipients into ten five-year clusters—1961–65, 1966–70, 1971–75, and so on to 2006–10—breaking out five major disciplinary fields (humanities, social sciences, physical sciences, life sciences, and engineering) and, within the humanities, three SED subfields (history, foreign languages and literatures, and American and English literature). Progress to the degree lengthened across all the arts and sciences disciplines during the 1970s and 1980s. But it lengthened most for graduates in the humanities. Graduates who received humanities PhDs between 1986 and 1990 spent 2.5 years more in graduate school than humanities graduates who completed degrees two decades earlier, between 1966 and 1970. And humanities PhDs who received degrees between 1986 and 1990 spent over four years more in graduate school than their 1986–90 counterparts in the physical sciences, three years more than those in the life sciences, and two years more than those in the social sciences.

Fig. 1
Fig1_time to degree

As of 2012, the most recent year for which SED data are available, the gap between the humanities and other disciplines is smaller but remains substantial: over a year longer for humanities PhD recipients than for PhD recipients in the social sciences and over two years longer than for graduates in the life sciences and the physical sciences (fig. 2).

Fig. 2
Fig2_time to degree

How one views these data on time to degree will vary, depending on what one thinks the data reveal about the risks and opportunity costs of attempting to enter the academic profession. Do we think time is an abundant or a scarce resource for students in their twenties and early thirties? Do we think it unreasonable to expect graduate students to invest a decade of a working lifetime and life span to earn a PhD?

In retrospect, we can see how the lengthening path to the PhD documented in SED data from the 1970s and 1980s coincided with concerns voiced at the time about the way a depressed academic job market was deforming graduate education, placing increasing pressures on graduate students for professional activity, especially publication, that up until then had been demanded only later. John Guillory captured the contemporary sense of what was happening in his 1996 essay “Preprofessionalism: What Graduate Students Want”:

What the market demands, incredibly, is a graduate student who is already in some sense a successful professional before that student can be considered for a position as a professor. In such a context, “professional desire” is contorted into the form of prematurity, of desiring something now—professional success—that can only be had later. This prematurity is phantasmic: it telescopes professional careers into the time period of graduate school and conflates graduate education with self-marketing, as though getting a job were somehow the culmination of a successful career. (4)

That preprofessional pressures for graduate students to publish would operate to lengthen time to degree I take as axiomatic. The same period, 1970 to 1990, was also notable for a rising bar for tenure and promotion and increased demands for publication that motivated a transformation of the tenure-track assistant professor position, especially in doctorate-granting research universities. The work and life of faculty members on probationary appointments changed from what had been a distinctly junior status quite sharply differentiated from the privileged standing tenured professors enjoyed. As the job crisis in the 1970s made assistant professor positions scarce, it also made them more privileged and protected, more like than unlike the senior professorial ranks in the conditions of their work. It became common for university departments with graduate programs to limit the exposure of their assistant professors to committee and other service obligations. Assistant professors came to have the same (lighter) teaching load as their tenured seniors, the same or sometimes even greater access to research leave, and the same exemption from teaching the required composition course and other introductory courses.

For actors on all sides—whether graduate students or faculty members in their dual roles as advisers to graduate students and members of search committees making hiring decisions—the sense that the bar for earning tenure was getting higher likewise worked to raise the bar for entry to a tenure-track position. To be competitive, graduate students afforded themselves the time to make the entry ticket of the dissertation as close as possible to a monograph manuscript ready for the publisher.

The downward trend in time to degree since 1990 is every bit as interesting as the prior upward trend—and may be more of a challenge to understand. Certainly, around 1990 time to degree and the closely related matters of attrition and completion rates emerged as prominent topics in the policy discussion around the economics of doctoral education. In Pursuit of the PhD, William G. Bowen and Neil L. Rudenstine’s 1992 follow-up to the 1989 Prospects for Faculty in the Arts and Sciences, by Bowen and Julie Ann Sosa, brought renewed attention to the economic issues in doctoral education—of high costs to universities, high opportunity costs to students, and questions about the investments (and returns on investment) of fellowship programs like those sponsored by the Danforth Foundation and the Woodrow Wilson National Fellowship Foundation. The Graduate Education Initiative that the Andrew W. Mellon Foundation undertook through the 1990s explicitly made shorter time to degree a prime objective and called attention to long time to degree and high noncompletion rates as problems graduate education needed to address (Ehrenberg, Zuckerman, Groen, and Brucker). More recently, the Council of Graduate Schools’ PhD Completion Project reflects and amplifies the concerns administrators and funders have expressed about the large investments of time and resources doctoral study has typically demanded from both the students who undertake it and the institutions that sponsor it. The data suggest that these discussions and initiatives may have had an effect.

A median summarizes a distribution with an equal number of values higher and lower than the single median number. As part of the MLA’s request for historical data on time to degree, the office of research asked NORC to develop frequency distributions detailing the number of graduates who took differing amounts of time, from six years or less to more than thirteen years. The distributions behind the medians turn out to be illuminating. As shown in figure 3 and figure 4 (in which the distributions have been aggregated to show time to degree in five groupings from six years or less to more than twelve years), the medians mask the real locus of action over the past fifty years, which has been at the extremes of shortest and longest times spent in graduate school. In the first three five-year temporal groups—1961–65, 1966–70, and 1971–75—a quarter of humanities graduates took six years or less to complete their degree programs—about the same portion that took more than twelve years. From the 1980s forward, the segment taking six years or less dropped to 10%, while the segment taking more than twelve years expanded to almost 40%, falling back to just under 30% from 2000 on. Figure 4 shows the same information but groups degree recipients across the ten time periods together according to the number of years they spent in graduate school.

Fig. 3
Fig3_time to degree

Fig. 4
Fig4_time to degree

As figure 4 may reveal most clearly, over the fifty years the biggest changes were a pronounced drop after 1975 in the percentage of humanities graduates who took six years or less to complete their degrees and a corresponding increase in the percentage taking more than twelve years. Changes over time in the three middle groups were modest by comparison.

Interestingly, a pattern almost identical to the humanities is evident across the disciplines. The life sciences, physical sciences, and social sciences, however, start in the 1960s with far higher percentages of graduates finishing in six years or less and end in the 2000s with far lower percentages taking more than twelve years to complete their degrees (fig. 5 and fig. 6).

Fig. 5
Fig5_time to degree

Fig. 6
Fig6_time to degree

A review of the history and historical trends still leaves us to grapple with the question of direction for the future. Figure 7 focuses attention on the 2006–10 group of program graduates and compares the disciplines by the percentage of those that had the shortest and longest time in graduate school. The figure reveals the large and even startling disparity between the humanities and the other arts and sciences fields. The percentage of 2006–10 graduates completing degrees in six years or less (the green bars) is three to four times greater in the sciences, engineering, and social sciences than in the humanities. And the percentage of 2006–10 humanities graduates taking more than ten years to earn a PhD (the red bars) is more than double the percentage in the sciences and engineering and 1.5 times the percentage in the social sciences.

Fig. 7
Fig7v3_time to degree

Circumstances where, whether by choice or necessity, 40% of a field’s PhD recipients end up taking more than a decade to earn their degrees seem unsustainable. Shrinking the surprisingly large group of degree recipients in language and literature who take an inordinately long time seems imperative. On the other hand, when thinking about practical measures to respond to this imperative, we need to recognize how the data hide individual histories, with their complex mix of circumstances. There is much we would wish to know and understand that these data do not and cannot tell us. As students pass beyond a fourth or fifth year, how do they support their graduate studies? If they are self-funding five or six years or more of doctoral education, how is lengthy time to degree intersecting with the troubled state of academic labor and the adjunct academic workforce? The data do, nonetheless, remind us forcibly why it is important for local programs to keep track of doctoral candidates and their progress to the degree.

David Laurence

Works Cited
Bowen, William G., and Neil L. Rudenstine. In Pursuit of the PhD. Princeton: Princeton UP, 1992. Print.

Bowen, William G., and Julie Ann Sosa. Prospects for Faculty in the Arts and Sciences: A Study of Factors Affecting Demand and Supply, 1987 to 2012. Princeton: Princeton UP, 1989. Print.

Doctorate Recipients from U.S. Universities, 2012. National Science Foundation. Natl. Science Foundation, Jan. 2014. Web. 9 May 2014.

Ehrenberg, Ronald G., Harriet Zuckerman, Jeffrey A. Groen, and Sharon M. Brucker. Educating Scholars: Doctoral Education in the Humanities. Princeton: Princeton UP, 2009. Print.

Guillory, John. “Preprofessionalism: What Graduate Students Want.” ADE Bulletin 113 (1996): 4–8. Web. 9 May 2014.

4 thoughts on “Opportunity Costs of the PhD: The Problem of Time to Degree

  1. Pingback: Important New Report on PhDs and the Time to Degree | Laurie N. Taylor

  2. J Lawrence Mitchell

    Another incisive report by David Laurence! Clearly, average time-to-degree isn’t that helpful in showing us what has happened over the years. We now need to broach some of the truly difficult questions.

    Reply
  3. Raghuveer Parthasarathy

    This is an excellent post. One comment: I’m skeptical that the giant spike in the humanities time-to-degree in the late 1980s (10.7 years!) in Figure 1 is real — I’ve been looking through this data, and it looks like data in the 2012 “survey of earned doctorates” report has values that conflict with earlier reports. I’ve put some graphs, and a link to tabulated data with sources, here:
    https://eighteenthelephant.wordpress.com/2014/12/04/universality-scaling-and-time-to-ph-d/

    Reply
  4. Pingback: The Personal Is Philosophical: Reflections on Reading 245 Connected Academics Proseminar Applications | Connected Academics

Leave a Reply

Your e-mail address will not be published. Required fields are marked *.