National Science Board Convocation On Graduate and Postdoctoral Education
October 9, 1997
My textbook being widely used, what that means is I still drag it out about once a month and have a look at it, and it does bring me sort of back home. I am delighted to be in Houston. I spent many happy and soggy years in this town. A day like today or yesterday is quite familiar to me. I want you all to know, as the Chancellor said, this is not usual Houston, but it is not uncommon. And when this would happen we would get our inflatable raft and we would put the kids and the dogs in it and we would go visit our neighbors up and down the street, because that was the only way you could really transport; our kids learned to swim that way, the dog learned to swim. That was life in Houston.
I am particularly pleased to be on this campus. I spent most of my years in a little campus to the west of here. The University of Houston has always been extremely important to the future of this city and to the State and it remains so today. It is always a delight to see what tremendous progress has been made since the last time I was here. I have many good colleagues on this campus and feel very close to the University of Houston, so we are delighted to be here today.
Let me say just a few things about this issue that is so vital to the country and is our focus. Suppose I were to construct the following argument: The first premise is that the Federal government is an appropriate source of support for fundamental research conducted at colleges and universities. Second premise: these investments contribute directly to education at the graduate and postdoctoral levels, as well as at the undergraduate levels. Therefore, it is appropriate for the Federal government to establish policy for advanced education in the sciences and engineering.
Well, I think it is safe to say maybe we could get acceptance of the first two premises, but the conclusion I think might give us more heartburn than a bowl of Texas jalapeno chili, which I encourage you to dig into while you are in town.
Today, I intend to examine the tradition and trends that have led us to this unsettling predicament. My remarks are entitled The NSF Role in Graduate and Postdoctoral Education: Will Traditions Survive the Tests of Time?.
Well, NSF, like other Federal R&D agencies, has a long-standing tradition of supporting the advanced education of students in science in engineering. It has served the nation very well -- in effect enabling research investments to do double-duty. They simultaneously advance the frontiers of knowledge and they provide the nation with a world-leading science and engineering workforce. Our system is the envy of the world for good reason.
Our system, however, is now facing new tests -- tests of accountability brought on by such factors as the Government Performance and Results Act, and the tests of practicality brought on by an uncertain labor market. These of course come on top of the test of survival brought on by a constrained budget environment. Reconciling our traditions with these tests of time will not be easy. In the end, we are compelled -- indeed I think required -- to take a thorough and fresh look at the underlying premises and assumptions which shape the Federal role in graduate and postdoctoral education.
This tradition, as I call it, did not emerge by happenstance or evolve slowly through Federal policies and programs, as you have already heard from previous remarks. Indeed, if we are to begin at the beginning -- which we might do appropriately so for science policy by referring to Vannevar Bush's Science: The Endless Frontier, which Dr. Kelly talked about, we can gain some insight into the issues that we face today by reflecting then on what Bush had in mind. His famous "Program for Action" explicitly linked research and advanced education, as Eamon [Kelly] highlighted. To quote, "The Government should accept new responsibilities for promoting the flow of new scientific knowledge and the development of scientific talent in our youth."
Bush in this way sounded a call for government investment jointly in fundamental research and advanced science and engineering education. It is noteworthy, however, that the prevailing view on graduate education in Bush's day differed substantially from what has come to pass. In his chapter on the "Renewal of Our Scientific Talent", Bush wrote: "...the road to careers in science and engineering is long and expensive; it extends at least six years beyond high school." Well, I think Bush may have been one of sciences great visionaries, but he really kind of missed the timeframe in which we are doing all this. He did not, I think, foresee that graduate doctoral education is in fact six or more years beyond the baccalaureate degree, and students are known to spend ten years, twelve years, longer in the university environment still thinking of themselves, and in some cases still being treated, as students.
Now, other documents from the Foundation's earliest days also shed some light on the agency's role in graduate advanced education. Graduate fellowships, for example, have been the mainstay of NSF since its founding.
Section 3 of the NSF Act of 1950, which I urge you all to get out and read at your bedside, authorizes and directs the agency to provide graduate fellowships for study and research in the sciences and engineering. This was the second authority specified in the original Act, followed only by "initiate and support basic research." This tells us that a direct Federal role in graduate education was foremost on the minds of Congress when the Foundation was created.
In fact, the first operational budget for the Foundation -- for fiscal year 1952 -- devoted $1.5 million to graduate fellowships, and only $1.1 million to research grants. That is why Terry Porter and others at NSF in our Division of Graduate Education keep talking about the good old days when they had the balance right.
Other snapshots from the historical record make clear that the Foundation's leaders appreciated the interplay between research grants and graduate education. The pre-Sputnik Foundation had difficulty securing even the most meager budgets, and the long-term nature of basic research was proving to be a tough sell in the Congress. Still true, to some extent, today.
It was in this spirit that then-NSB Chair Chester Barnard posed a rhetorical question in the Foundation's 1953 Annual Report. He asked, "What are the relatively immediate consequences of basic research?" And he answered his question with a three-item list. Items two and three were new knowledge and its potential uses - no surprises there. At the top of his list was: "the development of scientists [period]." The succinctness of Barnard's statement shows that the Foundation, then as now, recognized that the natural linkages between research activities and graduate education provide a compelling justification for the overall investment. In those intervening years, engineering was born in NSF took on a major responsibility so we now would replace by science and engineering.
The NSF Annual Report for the preceding fiscal year, 1952, emphasized the same point, and it did so in a way that foretold one of the most difficult issues we face today. It contained a short paragraph under the heading, "Research Support Strengthens Teaching." And the paragraph began:
It then adds: "At the same time, the research grants provide added funds for research assistantships and for materials and equipment for research."
Well, this final sentence is significant for two reasons. First, it directly links the Research Assistantship with education. And second is the fact that it places RA's, research assistantships, on an equal level - at least grammatically - with machinery and equipment. We have since cleaned up or at least changed our grammar, though our administrative and accounting systems still leave something to be desired. And I will return to that point.
From this brief review of NSF's early history, the basic outlines of our role begin to take shape. Advanced education in science and engineering has always been at the core of NSF's mission and programming. From its very first actions, the Foundation assumed both a direct role in graduate education by supporting fellowships and an indirect role by explicitly recognizing the contribution of research grants to education.
Over the years we have seen this basic outline set the stage for a highly productive set of investments. It is now October. The Nobel prize announcements are coming out day by day. We should not be surprised to see former NSF graduate fellows garner more than their share of awards. Over the years, they have accumulated at least 15 Nobel Prizes, and scores of virtually every other top award offered in science and engineering, including the National Medals of Science and Technology. Rumor has it that at least one former NSF fellow went on to run the agency, but it is unclear how that is viewed in the eyes of the program.
The RA mechanism also brings a host of educational advantages. We have all known students working in laboratories supported by grants from two or more different agencies. One grant could be from NSF and others from mission agencies. The student therefore learns first hand that consideration of use often move in tandem with one's quest to extend the frontiers of knowledge. There is no substitute for the insights gained from such an experience.
Moreover, the new knowledge and technologies that flow from the research performed by the student contribute, in a most efficient manner, both to advancing the frontiers of science and engineering and to mission needs of one or another agency. When industry is also a partner, economic benefits can flow as well -- in keeping with the notion of technology transfer being a contact sport.
Until very recently, we were inclined to point to success stories like these and relax in the luxury of knowing that our system was working on the nation's behalf. Unfortunately, our traditions are not holding up so well against the tests of time. For starters, recent graduates have told us in no uncertain terms that their advanced education is out of alignment with the realities of today's job market. The overall health of the economy has improved the situation somewhat, but there are still causes for concern.
Unemployment of PhDs has never been very high, and the latest data suggest that it is returning to its historically low levels. Unemployment of physics PhDs, for example, reached 5.3 percent in 1993, quite high compared to the tradition of the field. But it was back down to 2.9 percent by 1995, according to the latest results from NSF's Survey of Doctoral Recipients. The overall unemployment rate for science and engineering PhDs remains under 2 percent.
Numerous studies and reports -- notably the 1995 COSEPUP report , The National Academy, and the report of the NSB task force -- have addressed these issues and trends. This has prompted efforts like NSF's IGERT -- the Integrative Graduate Education and Research Training Program -- which we believe will give institutions a means to test innovative approaches to graduate education.
In addition to the uncertain labor market, NSF's investments in advanced education are also being put to the test by everyone's favorite 800 pound gorilla, inside the 'Beltway' -- the Government Performance and Results Act. For short, it is called the Results Act. The Results Act brings a new dimension to this discussion. If we all agree -- and I think we do -- that NSF's role in graduate and postdoctoral education is appropriate and beneficial to the nation, then we face the requirement of evaluating the return on this investment in a systematic way. The rules of the game have changed. The purpose of the Results Act is to see if we have been successful in carrying out our mission.
One of the five outcome goals specified in our Results Act strategic plan bears directly on our role in graduate education. It reads: "A diverse, globally-oriented workforce of scientists and engineers." Well, as this wording implies, we are seeking to foster a science and engineering workforce that reflects the full diversity of our society and is able to compete successfully in today's global labor market. But documenting success on these scores is not a trivial task.
For direct support mechanisms -- such as fellowships and traineeships -- we are on relatively solid ground. We know who has received our support, we know how diverse a group it is, we know how long it took them to earn their degrees, and where their careers have led them. The Results Act poses some interesting hurdles - defining quote/unquote "successful outcomes" being one of them. Does success mean landing a tenure track position? Is someone who becomes a high school teacher or an agency head a failure? These are difficult questions to address in the context of the Government Performance and Results Act, but at least we have the confidence of approaching them with a reliable data set.
For research assistantships, however, we are on somewhat shaky ground in terms of making the same kind of determination of success. First of all, the funds are controlled by the institution, and the PI and the department have the final say on who gets funded under particular grants. We do not always know which students get support. We have little or no information about the demographic characteristics of the students being supported, which makes the RA a weak mechanism for addressing issues related to under-representation. It is even difficult for us to get an accurate count of how many students receive our support. Different institutions have different databases, formats, policies, and procedures. And, sometimes NSF support is shown separately from other agencies, sometimes it is not.
The sources of support can change over the course of a year, as grants end and new funding sources replace them. There are also Privacy Act restrictions on the students' records, which may prevent us from obtaining basic information like their social security numbers.
In sum, there are a host of issues to address. Our Division of Science Resources Studies is currently working with professional societies and institutions on this, with the goal of developing a reliable and feasible means of identifying and tracking the students we support.
This does leave us with a disquieting quandary. We know that in a given year we support roughly 3,500 students through fellowships and traineeships. Our best estimate of the number of students that we support through research assistantships is around 11,000. That figure reflects the students who report that the RA was their primary source of support - meaning that they relied on it for more than 50 percent of their financial income. If we count all the students working as research assistants on NSF-supported projects, the number supported approaches 17,000.
And while the totals are somewhat elusive, these figures tell us that we support 3 to 5 times as many students through indirect and perhaps un-measurable means as through fellowships and other direct mechanisms. I can recall Shirley Malcom reminding us when we talked about the Results Act that the real test under the Results Act is "to measure what we value, not to value what we can measure." And that is proving to be unexpectedly challenging task.
I should add that these figures are a reasonable reflection of the government as a whole. NSF supports only around 20 percent of the graduate students and postdocs in science and engineering who list Federal government sources as their primary means of support. And those students in turn are only about 20 percent of the total number of science and engineering graduate students in the country. When we look across the entire Federal government, we see a similar mix of support mechanisms. For students whose primary source of support is the Federal government, nearly two-thirds are supported through research assistantships, while barely one-fourth receive fellowships and traineeships.
Again, with such a large component of the Federal investment in advanced education occurring indirectly, that is through research grants, holding ourselves accountable under the Results Act poses a potentially insurmountable challenge. It is a real test of time for our long-standing tradition.
All of this brings us face to face with questions that have percolated and persisted since the Foundation's creation -- if not longer.
What is the purpose of graduate education? Who should fund it? By what mechanisms? Should we change the mix of mechanisms? How should educational institutions and funding agencies share responsibility? What is the institution's responsibility to students after graduation? What is the funding agency's responsibility? How can we incorporate expectations of future employers? Should we incorporate expectations of future employers? How do we incorporate measures of success and failure into our policies and programs? Finally, who should have the lead on policy issues? Funding agencies or Institutions, or should that be shared?
As I said at the outset, questions like these may give us some heartburn, so that is why it is good to be meeting here in Texas, where the phrase the "the hotter the better" applies to more than just Texas chili. So with that in mind, I suggest that it is time for us to dig in. And I look forward to hearing from my colleagues.
Thank you very much.