Stanford University News Service



CONTACT: Stanford University News Service (650) 723-2558

Doctorate surplus in science, engineering is ongoing, researchers say

STANFORD -- Universities in the United States are producing about 25 percent more doctorates in science and engineering fields than the U.S. economy can absorb, according to a new study by researchers at the Rand Corp. and Stanford's Institute for Higher Education Research.

While some academic and science organizations have argued that government should expand research funding to avoid wasting these resources invested in doctoral training, sponsored research actually increases the long-term overproduction of doctorates, say Stanford Professor William Massy and Charles Goldman of Rand, in a study done with graduate students Marc Chun and Beryle Hsiao.

Economic models prepared by the group show that "you can increase sponsored research by 10 percent, and while you are in the process of making that increase -- let's say you move it up at 2 percent a year for five years -- you surely will sop up the unemployment," Massy said during a recent interview.

"The whole system will be expanding and people will get the kinds of jobs they were trained for. However, as soon as you stop increasing it and go back to a steady state -- not a decrease but just stop increasing research funding -- all of a sudden the underemployment comes back. In fact, it comes back worse than it was before because the whole system has scaled up."

In an analysis of 13 science and engineering fields covering 210 doctorate-granting institutions and more than 1,000 educational institutions that employ people with doctorates, the researchers found that supply and demand do not work in the usual way to regulate the employment market. In labor markets, when job opportunities decrease fewer people usually seek to enter the field in response to the reduced opportunities. In the case of Ph.D.s, however, the researchers said, they found that "neither departments nor prospective doctoral students take close accounting of the doctorate employment gap."

Interviews with 300 faculty members on 19 campuses indicated that doctoral admission decisions "are driven not by the output market [for doctoral degree holders] but by the academic department's own production needs" for teaching and research assistants, Massy said. Teaching and research assistantships are temporary, part-time jobs for doctoral students that are normally thought of as the byproduct of producing doctorates. "It's kind of the tail wagging the dog," he said.

The only way to solve the long-term underemployment of doctoral degree holders in sciences and engineering, the researchers say, is for academic departments either to reduce the number of doctoral students they admit or to convince more potential Ph.D. candidates not to seek the degree.

Both are difficult to do, Massy said.

Faculty and administrators who make admissions decisions tend to admit their targeted number of doctoral students, regardless of changes in quality of applicants, he said. "They do that because they must have the Ph.D. students for teaching assistants, for research assistants and because faculty have a sense, in certain places, that they really need Ph.D.s to keep intellectually alive."

At his own School of Education, for example, Massy said, "every faculty member believes that he or she should be able to admit one Ph.D. student a year. Departments and schools have different numbers, but there is a sense that this is what we are entitled to as faculty. It's part of our intellectual culture."

The targets for Ph.D. admissions vary greatly with the type of institution and by field, he said. In electrical engineering, the volume of sponsored research grants tends to drive the number of doctoral students needed to do the research, and that field has the highest number of doctoral students per faculty member at the institutions studied. In other fields, such as mathematics, chemistry and economics, undergraduate enrollment is a greater influence on the number of doctoral students admitted because these departments are responsible for teaching many undergraduate general education courses, and the departments admit doctoral students to help teach those courses, Massy said.

"To put it bluntly, in departments where general enrollment is a problem, at least some kinds of institutions will have to change their mix to use more faculty and fewer TAs [teaching assistants] to teach undergraduate courses, because that has a dual effect: It reduces the number of doctoral students produced and it increases the demand for faculty," thereby creating higher numbers of permanent jobs for the doctoral students who are admitted.

The best place for this change to occur would be at the lowest-ranking doctoral-training departments, but faculties in these departments are not likely to make the change voluntarily, he said, because having fewer graduate students would tend to have a negative effect on their own careers. It is more difficult for them to publish new research results without new graduate students and they most likely would have to spend more of their time teaching undergraduate courses than the advanced courses many prefer. "It goes against a whole lot of forces," Massy said.

Overproduction may cease in time, he said, because of rising pressure on institutions of higher education to increase the quality of undergraduate education. Some parents and students complain that undergraduates are too frequently taught by doctoral students rather than faculty. "I think that is going to push [some] institutions toward enriching the faculty mix and decreasing the doctoral students" who teach undergraduate classes, Massy said.

Potential doctoral students also might apply in smaller numbers if they have better information on their permanent job prospects, he said. In the absence of data, they may be too influenced by the success of their primary role models -- faculty members with their own research laboratories. "If they only knew that a small fraction of the people who start off ever get there, it might make a difference at the margin," Massy said.

Problems in providing such information result from the difficulties encountered in tallying the numbers of graduates who are actually employed after graduation in jobs that use their degrees. Some data count temporary post-doctoral employment or "nomadic" employment as a lecturer as if the graduate had reached his or her goal, he said. Other data sets merely count the total graduates who are employed. Doctoral students are likely to have skills that will lead them to compete better than the average applicant for jobs, he said, even if those jobs are as taxi drivers.

Using the best data available, even the Rand and Stanford researchers were not able to come up with precise figures on the underemployment for science and engineering doctoral graduates. Their best estimate is that over the long term and across all science and engineering fields, three of four doctoral degree holders eventually get jobs related to their degree qualifications. The actual percentage varies by field, is dependent upon assumptions about the proportion of doctoral students from foreign countries who fill some of the jobs available in the United States, and can change temporarily when shortages occur in a given field. The estimate also does not count any jobs that U.S. citizens with doctoral degrees might hold abroad.

"For the five engineering fields [studied], the employment runs between 25 and 50 percent of Ph.D. production under the assumption that half of the visa-holding graduates remain in the United States," the researchers report. "The employment gap remains positive, between 5 and 20 percent of total degrees, even if we assume that none of the foreign graduates remain. . . . If we want to characterize the gap for a lay audience, we would choose 25 percent of total degrees -- the low end of the 50 percent foreign retention range. For the five engineering fields, this would imply that about 200 new Ph.D.s, on average per field, would fail to find suitable doctorate-level employment."

The 25 percent who do not make it usually do get jobs for which they may be overqualified. "You have valuable training, so you probably will do fairly well in your life," Massy said, of doctoral degree holders. "That's another reason for saying information will only affect some people at the margin, but not most."



This is an archived release.

This release is not available in any other form. Images mentioned in this release are not available online.
Stanford News Service has an extensive library of images, some of which may be available to you online. Direct your request by EMail to

© Stanford University. All Rights Reserved. Stanford, CA 94305. (650) 723-2300.