What's wrong with graduate education these days? Is it what Ph.D. students are missing -- teaching skills perhaps, or the ability to branch out beyond their area of specialization? Or is that too many students take too long to finish?
At a discussion Wednesday as the annual meeting of the American Political Science Association kicked off in Boston, panelists agreed that the answer is "all of the above." And they also agreed that there may be a contradiction -- a "collision course," in the words of one expert -- between the desire to add more elements to graduate education and the increasingly shared view that universities need to cut "time to degree." While the discussion focused on political science, those on the panel have also been involved in graduate education generally, and said many of the issues apply elsewhere in the social sciences.
So what's missing in the training of new Ph.D.'s?
"Today’s graduate students have a degree of training and proficiency far in excess of those I had when I started out," said Graham Wilson, a professor of political science at Boston University who earned his Ph.D. in 1975. "But what is happening to wider knowledge? We want to hire [new Ph.D.'s] to teach a number of different courses across the program or to be part of liberal arts programs or to be in interdisciplinary program," he said.
Wilson, who was a graduate director at the University of Wisconsin at Madison earlier in his career and is about to take up that job in his department at BU, said that "we've seen over the decades a progressive narrowing" of the number of fields in political science that graduate students are expected to study in depth. However "admirable" much of the work being produced by new Ph.D.'s may be, he said, "we are producing a narrower product" that doesn't fit with what hiring departments want. And while professors like to talk about grand philosophical issues when considering what a new Ph.D. should learn, he said, it's important to remember that graduate education "is very much about getting jobs."
Theda Skocpol, a professor of government and sociology at Harvard University, said that these issues aren't new. More than five years ago, a panel appointed by Skocpol when she was president-elect of the political science association issued a report  calling for more breadth in graduate training. The problem, she said, is that departments haven't fully embraced the ideas -- and the field has continued to grow.
“We do risk overspecializing and it’s a conundrum. There is more to learn in every subfield,” she said. "But I think it behooves us to recognize that on the job market, whether it is going off to work in a college where they are going to want the student to teach multiple courses, or having to develop research ideas that will engage the public over the course of a career, students have to have the ability to combine approaches and address real world questions." With so much specialized knowledge, she said, "it's too easy for us to wander off into very specialized communities and I don't think that will be in our interest."
Political science is also a field in which there has long been tension between subject matter and methods -- and some said that may contribute to the problem. Skocpol said it was important that new Ph.D.'s be well versed in more than one methodology.
"There’s the constant struggles of substance vs. methods," said Elliot E. Slotnick, associate dean of the Graduate School at Ohio State University. He said he believes that many departments have "gone too far" in favor of methodology and formal theory, and that he worried about "how easy it can be to have someone teach a course on legislatures without really knowing anything about Congress."
Teaching generally is another area where the panel said graduate programs need to do much more. Given that most students will end up at institutions more oriented toward teaching than their Ph.D.-granting universities, graduate education has been remiss in its emphasis on research, the panel said.
So if graduate programs should offer broader training, and also teach their Ph.D. students how to teach, should they stop worrying about "time to degree"? Absolutely not, said the panelists.
Skocpol, who recently completed a term as dean of Harvard's graduate school, is a "time to degree" hawk who aims her firepower not at graduate students but their departments. At Harvard, she instituted an unusually successful reform  in which for every five graduate students in years eight or higher of a department's Ph.D. program, the department loses one admissions slot for a new doctoral student. The program has had an immediate impact, with many departments that would have lost admissions slots moving quickly to get their long-time grad students through. If departments aren't pressured to work with grad students, she said, they will tolerate students lingering in programs.
"It's a terrible waste of institutional money and of people’s commitments in life” for people to be Ph.D. students 10 years and more, she said.
In fact, while Skocpol pushed for better stipends for graduate students, she said that too much money can discourage grad students from finishing up. She advocates linking funds to timetables.
At Ohio State, which as a large public university operates in a very different financial reality than that of Harvard, Slotnick said that speeding up completion has also been a top priority. The university, for example, links some fellowship support for dissertation writing to the time that a person has taken to get there -- so there is an incentive to reach the dissertation-writing stage at a good pace.
Further, he said, the university is starting a "continuous registration" requirement under which anyone wanting to remain a graduate student must pay a modest fee (a few hundred dollars). This should help departments identify who is out there, theoretically finishing a program -- and may prod some grad students to take their completion plans off the back burner.
Part of the problem at a place like Ohio State, he said, is that departments don't even know the status of all of their graduate students.
So how could departments and grad schools shift their programs while not extending their duration? The panel offered several ideas.
Redefine time to degree. Skocpol said she is dubious of many of the statistics because they are inconsistent in how they count and they play down the costs of students who drop out. She devised a different standard -- counting every graduate student who entered a program and tracking the student for the next 10 years. Those still enrolled at the end of 10 years get a 10, those who have earned their Ph.D.'s earlier get a number representing the years it took them, and those who dropped out get a number reflecting how many years they were enrolled. Then Skocpol divides the total by the number of doctorates awarded.
The result, she said, is the "human years invested per Ph.D." When she did it at Harvard's graduate school, she found departments ranged from 4 to 16 years -- with the higher figures representing not only more years of work on a doctorate, but more work prior to dropping out. These figures, she said, "are embarrassing" to departments -- and can be used to promote discussion and reforms.
Make better use of summers -- including the pre-enrollment summer. Many graduate students need extra math assistance or (for those from some parts of the world) English assistance, and this should be taking place the summer prior to grad school, not during the program itself. This allows more students to move at a good pace from day one. Further, they said that summers are ideal time for interdisciplinary programs involving graduate students from multiple departments or from different parts of a discipline. While noting that summer programs require funding for grad students, the general sense was that summers are not sufficiently utilized, especially early in a Ph.D.
Build ties to teaching institutions. Slotnick said that Ohio State has used the Preparing Future Faculty  program -- an initiative of the Council of Graduate Schools -- to offer a range of programs for graduate students about what their future jobs are likely to be. A key lesson, he said, was the need to get graduate students on teaching-oriented campuses so they could see what such careers look like. Ohio State works with faculty members at Kenyon College and other liberal arts colleges in the state -- with programs at Columbus and the various college campuses. Some graduate students have mentors at the colleges and get to spend time there and teach -- while they are still shaping their graduate program. In this way, an awareness of what potential employers would want is clear much earlier than a job talk.
Discourage early publishing and presenting. Slotnick said that in terms of grad students having more time for breadth, they can cut back on the trend of trying to publish articles and make presentations early in their graduate student careers. He said that many of these papers just aren't that good -- and hurt the reputations of the students and their programs.
Create "research workshops." Skocpol said that the social sciences need to embrace the positive role of laboratories in the physical and biological sciences by encouraging more team work and sharing of work. In the workshops, graduate students present papers for critique by colleagues -- and get into the regular pattern of presenting work, responding to critiques, and sharing ideas. Even if participants are from the same discipline, they will not be as close to a project as the student him or herself and the dissertation committee. As a result, the grad students learn to respond to people without as much knowledge as they have. This not only encourages a broader perspective, but prepares student for the job hunt. "When you go on the job market, it's not the talk that gets you the job, but how you answer the questions," Skocpol said.
Teach time management. Skocpol said that faculty mentors should actively teach time management, especially with regard to teaching undergraduates. "Teaching them how to be good teachers and how not to let it eat up all of your time” is key, she said. Graduate students need to be told that they can be responsive to undergraduates without "being available 24 hours a day," and the need to learn that if a teaching assistant position is supposed to represent one fifth of their workload "how to do it in one fifth of their time."