NIST logo
*

Testimony of James C. McGroddy
Retired Senior Vice President, Research, IBM Corporation
Committee on Science
U.S. House of Representatives

March 11, 1998

Thank you, Mr. Chairman.

My name is James C. McGroddy, and I am a retired senior vice president, research, of the IBM corporation. I was employed at IBM for over thirty years, most of this in its Research organization. From 1989 to the beginning of 1996 I was responsible for the worldwide IBM Research organization, consisting of over 2500 technical professionals in seven laboratories around the world. I retired from IBM at the end of 1996, and my current focus is public service activities, the majority of which are with the US government and universities. Among other activities, I am currently the chairperson of the Visiting Committee on Advanced Technology at the National Institute of Standards and Technology, a member of the National Research Council's Government-University Industry Research Roundtable, and a member of a number of university "visiting committees".

I would like to thank the Committee for inviting me to testify before these important hearings, and I hope that my discussion of some of what I have learned both in industry and through contact with government laboratories and universities about the effective management of scientific research programs will assist the Committee in its important work.

The question of the degree of management which is appropriate for the scientific enterprise is one which is currently the subject of much debate. By management, I mean a set of processes external to the individual scientist, although usually involving him or her deeply, which affect the allocation at various levels of resources, the prioritization of fields and projects, the evaluation of progress and results, and the evolution of agendas, institutions and individuals involved in the scientific process. In my view, clear thinking in this area is often inhibited by much of the current terminology. Terms such as "unfettered research", "curiosity-driven science" and "strategic research" are but a few that come to mind. In what follows I hope to share what I have learned over my career as a scientist and a manager of scientists about the management of science and the scientific enterprise .

Over fifty years ago the rationale for major federal funding of science was set out with great clarity in Vannevar Bush's report to President Truman, "Science, the Endless Frontier". This report motivated the public funding of science by its demonstrated and projected huge positive impact on a strong national defense capability, enhanced health of the citizenry, and national economic performance resulting in high quality jobs. The payoff has been and continues to be enormous. Science has provided much of the root structure for technological revolution in many fields, and has delivered on the other goals as well. I know of no serious student of history who would today substantively revise Bush's rationale or conclusions in any major way, other than perhaps to add a fourth area of impact, the improvement of our management of our environment.

Given that there is no question that science, funded by the federal government as well as by others, has proven of enormous benefit to the nation, one can reasonably ask about the positive or negative effects produced by the various mechanisms in place by which science is managed. Peer review, a mechanism explicitly internal to the scientific community, has clearly been and continues to be a major element of the success of science itself, science in its own terms. But peer review is not in itself sufficient. I believe that the history of the scientific roots of the technological revolution, as represented for example by the microelectronics industry, the key enabler if the information revolution, will show that science has also benefited, both in the quality of science itself, and most certainly, in its ability to contribute to Bush's three goal areas, by a number of mechanisms which couple the science to its larger societal goals. When science is effectively managed, via a collaborative effort of the scientists themselves and their supporting and benefiting constituencies (or their surrogates), we get the best of both worlds.

My experience as a scientist, doing first relatively basic, then more applied, research, and as a manager of scientists, leads me to propose three principles which can be used to develop an effective system for the management of science. Such a system must deal with the allocation of resources at various levels, based on expected impact; it must contain a system for the measurement of progress and the evaluation of results achieved, both as science, and in terms of impact on society; and it must embed mechanisms for increasing the effectiveness of the investments made, both in the quality of results achieved and the pace at which they are achieved. The system which we evolved over a number of years during my tenure as head of Research at IBM appears to me to have accomplished these goals, building on the three principles discussed below. This management system is documented in detail in a 1995 National Research Council report entitled " Research Restructuring and Assessment : Can We Apply the Corporate Experience to Government Agencies ?" I have provided a number of reprints of the appropriate sections of this report for the Committee's use.

The three principles which I believe provide the basis for developing an effective management system for science, one which provides the necessary freedom and flexibility for work at the limits of the human imagination, as well as the stimulation and coupling which increase the likelihood and pace of resultant impact, are described below.

  1. The Management System must be Aimed at Maximizing Value. The people who do science, the scientists, are motivated by a variety of sources - curiosity, a desire to move the world forward, glory, financial rewards, etc. The sponsor, the provider of the resources, is usually motivated by a somewhat different set of objectives, including new product technologies, the education of students, or support of a particular economic sector such as energy. My first principle is that it is the job of both those who do, and those who manage, scientific ( or technological ) research to aim toward maximizing the Value of the research efforts, where Value is    defined in terms of the sponsor and the objectives of the larger institutional setting of the researcher. The researchers must initiate and drive the debate necessary to develop a clear and agreed definition of what constitutes Value. It is this understanding of Value, and of how it can be measured, with more or less precision, which is the foundation of any management system. Clearly what constitutes Value in a university research laboratory will be different from what constitutes Value in a NIST laboratory, and different from what constitutes Value in a particular industrial setting. Achieving agreement on the elements of Value provides the basis for measurement, particularly retrospectively, and this measurement will inevitably demand input from the downstream beneficiaries of the research.
  2. Effective Coupling Along the Path to Value Delivery is a Requirement. Science done in isolation is rarely the best science, and isolation radically decreases the likelihood, and the pace, at which the science will be exploited beyond the scientific community. The notion of research laboratories as monasteries on hilltops is long dead. As has been often pointed out, technology transfer is a contact sport. An effective management system will strongly encourage all its participants to develop links outside their own scientific sector, both because it is often the case that the most fruitful areas for discovery lie in the boundaries between traditional fields, and because awareness of the world of potential applicability downstream has often been shown to be an effective way to steer the science itself in more interesting and fruitful directions. Downstream awareness on the part of the research community, as well as upstream awareness by the technologists, are major accelerants of progress.
  3. The Management System Must Contain Mechanisms for Stopping Good Work. Many of the forces at work in the scientific community drive toward progress in a narrowly focused forward direction. Many of the mechanisms at work provide positive feedback, which encourages professors to drive their students to become clones of themselves, and for researchers to stay focused within a particular field of work. It takes a particular boldness on the part of an individual scientist to set off in a new and unexplored direction, to look for large nuggets of gold in areas where none have been found, rather than pan in established and crowded areas. Since resources are finite, and one wants to apply all the resources to efforts that are at least well into the "good" along a value scale, it follows that reprioritization will only take place by stopping "good" efforts, with the resources retargetted to efforts that promise to be "outstanding", far better than "good". Encouraging researchers to change what they are doing is the most difficult challenge, as pointed out long ago by Machiavelli. Yet without change driven both by the scientific community themselves and the management system, major opportunities for progress will be lost.


Science is not so different from other human activities that it cannot benefit from external inputs, from management. And science is too critical to our economic success for it to be shortchanged either in the level of resources we devote to it as a nation, or the wisdom with which we manage this critical resource, this large investment. Science, and its benefits, have become increasingly internationalized, and the world in which science is embedded has changed radically. We must as a nation ensure that, at an output level, we maintain an effective and broad scientific enterprise. We must get away from debates which focus only on inputs, which transmogrify progress in the pace of technological exploitation of science into death notices for basic research. And as we expand the resources which the nation devotes to the scientific enterprise, we must , both in the scientific community and as a nation, commit to continuously improving the quality and efficacy of that science. Creative and innovative management has made dramatic improvements in almost every sector of our society, and there is no reason to believe that similar progress cannot be made in science.

I found it revealing to read the most recent issue of the IEEE Proceedings, which celebrates the 50th anniversary of the invention of the transistor. This issue contains reprints of some of the original papers by Shockley and his team at Bell labs, as well as papers pointing to what might lie beyond semiconductors in providing the ten-times-every-five-years progress we have come to expect in this area, and which has contributed so much to our economy. What is very clear is that the transistor resulted from an extremely creatively managed mix of science and downstream objectives. The initial results were published in the world's most prestigious Physics journal, the Physical Review, and the Nobel Prize followed. I think that no-one would dare to argue that the science was diminished in value as science by the connection to a bold technological objective. In this same issue , there is a paper which builds on two recent Nobel-winning pieces of science, the scanning tunneling microscope and carbon buckeyballs, and points to what might become molecular-level transistors, continuing the remarkable progress of digital chip technology well into the next century. Science is truly the root system of technology, and the management of science is an important lever to enhance the connection of the roots to the tree.