Generating Science : Productivity and Policy

No one wants to become just another statistic, scientists least of all. We cherish our individuality, our march to a drummer whose beat most people never hear. And isn't our profession built on the unquantifiable flash of creative insight? Because statistical analysis concentrates only on average behavior, it is destined to miss outstanding individual achievement. Yet it is valuable to find the common threads that define our own special herd. What kind of analysis would help scientists understand themselves? One approach is to consider how we do science as we age. From children of the Great Depression to baby boomers, age and the common experiences that define a generation are deeply important. The shared events that form a scientific generation, such as the one that developed the atomic bomb, could equally well set the trajectory of a research career. The fact that some career elements are beyond any individual's control may be hard for self-motivated research types to accept, but is important for our profession to know. These thoughts came as I read Striking the Mother Lode in Science: The Importance of Age, Place, and Time (New York, Oxford University Press, 1992) a new book by Paula Stephan, a labor economist at Georgia State University, and Sharon Levin, an applied microeconomist at the University of Missouri, St. Louis. The book pulls together data about scientists, the science they make, and when in their careers they make it. It analyzes the influence on scientific productivity of age and cohort (the demographic term for those who simultaneously experience a crucial life stage). As most scientists would, the book defines the crucial event that begins a scientific career as the earning of a doctorate. The era in which a doctorate is earned, the authors propose, may play a crucial role in shaping a scientific career. Insights from the book and from a conversation I had with Stephan throw light on my own scientific track. Because the authors trace the past effect of funding policy in the United States on how we do science today, examining these statistics may help the U.S. government fund research more effectively in the future. Unlike other studies that examine selected high-achieving groups, such as Nobel laureates, Mother Lode looks at across-the-board data. It uses the "largest and most comprehensive longitudinal study of [Ph.D.] scientists in the United States," the Survey of Doctorate Recipients gathered by the National Research Council. Armed with this, and with the Science Citation Index (SCI) from the Philadelphia-based Institute for Scientific Information (ISI), the authors correlate productivity with age and cohort. As in many such studies, it is at the intersection of the quantitative and the qualitative that questions arise. The data set is large, the statistical analysis sophisticated--but how to define productivity? Stephan and Levin take an imperfect but defendable path, likely the only possible one: They count journal articles reported in SCI, the listing of research papers and their citations issued by ISI. To correct for coauthorship, they award each author a proportional piece of the article; to assess quality, they use the SCI impact factor, which ranks each journal by how often its publications are cited. Even corrected for authorship and journal (and the corrections carry their own share of troubling questions), a count of articles does not capture the complex nuances of how scientists produce science. As Stephan and Levin themselves recognize, mere counting lets slip through that stroke of understanding, the single idea that can change a science. Nevertheless, the study addresses what research in the aggregate is meant to do: produce new knowledge stored in verifiable form for later use by science and by society. In this sense, the book is a valid analysis of the most important endeavor of science as a profession. It weighs the research product. The measures may be imperfect, but the data are seductive. I couldn't resist comparing my publication record to the quoted results. (Few scientists could resist.) In my category, academic physicists, the average researcher turns out about two publications per two-year period, over careers spanning more than 30 years. The rates are different for earth scientists, biochemists, and physiologists, but not vastly so. More important yet is the finding in Mother Lode that the productivity of the average scientist decreases with age, in almost all the research areas and environments the book examines. In some specialties, the decline occurs over a whole career; in others, it follows a rising period, so there is a productivity peak. The book spells out the details, but the important thing is that it happens so widely. Given that scientists as a group are aging (in 1973, 25 percent of us were under 35 years old; today, only 10 percent are), the serious implication is that the profession is less productive than it once was, and will produce even less as its average age increases. (I emphasize again that "productivity" here means only the count of journal articles; the authors use no other measures of scientific achievement.) Is the decay in productivity an inevitable result of aging? Is it influenced by cohort effects, the special features that come from earning one's degree at some particular time? Some answers may come from the overlay of personal experience onto the statistics. This is why I want to talk about my own career, which has been heavily influenced by issues raised in the book. I belong to the very special generation in American science that was affected by the startling launch of the Soviet Sputnik satellite in 1957, when I was a college sophomore. The launch was not what turned me toward science, which I had chosen much earlier. But as the U.S. worked to match the Soviet triumph, the resulting public approval of science, the federal support for academic research and its personnel (growing at nearly 20 percent a year in inflation-adjusted dollars, notes Mother Lode), the prospect of a choice of jobs--all these kept me steadily on course to my doctorate in condensed-matter physics. After taking my degree in 1967, I turned to industrial research. But I had picked a company in which I could not do challenging research. It was when I looked for an academic job in 1968-69 that serious shock set in. Along with other Ph.D.'s driving cabs and waiting tables, I discovered that growth had ceased. The system wasn't following through on the promise it had held out. Fortunately, after much worry and effort, I found an academic position that has turned out well, at least as defined by Mother Lode. My productivity is more than four times the mean for my category, and is still rising. One moral of my story is that statistics do not describe the individual. Despite my start from a cohort that was set up by the promise of massive funding and was brought down when that funding contracted, my tale ends happily. But not entirely; there has been a price, consistent with conclusions in Mother Lode. The price is not a loss of the researchers' motivations. Based on studies cited in the book, these motivations are "ribbon," the search for recognition; "gold," the desire for money; and "puzzle," the ineffable pleasure of the scientific chase. Each still spurs me. What has changed is my tolerance for what must be done to seek ribbon, gold, puzzle--the incessant rattling of the tin cup needed to maintain a research effort. It is not that funding should be noncompetitive; it is that the search for dollars was already hard when I began, and has now become so consuming that it harms me, and our whole scientific enterprise. This has a lot to do with the period in which I got my Ph.D. I have long realized that colleagues five to 15 years older are the last group with some assurance of years of steady support from a single agency; my generation found money only with unending effort that rarely guaranteed continuity of research and researchers for more than a year or two at a time. This unstable way of life has, finally, created an air of anxiety that erodes the dignity and self-respect of scientists. For senior people like me, the anxiety is about keeping our self- determination, a large part of what brought us into science-- extremely important, but probably no longer career-threatening. For graduate students and postdoctoral researchers, it is a far deeper angst about even being allowed into the profession, to begin one's own quest for ribbon, gold, puzzle. A few young scientists will do very well, but any older mentor must weigh what it means to send yet more bodies into the competitive scrimmage. When I hear that scientists in their mid-30s are beginning their third postdoctoral appointments, still not independent researchers; when I see that academic openings elicit applications by the hundreds, I can only feel that something is deeply wrong, which diminishes the pleasures of research and the value of science for all of us. Are there any answers for American science? I asked Stephan what she would recommend if President Clinton were to call. The response came in a heartbeat: Find a way to give science funding a moderate but known rate of increase, year in and year out, rather than a Sputnik era-like 20 percent increase one year, followed by a 5 percent drop the next. Only this, she feels, would give scientists the "research horizon" necessary for good work. What "moderate" means is open to debate, and any policy that could be distorted to seem like a science "entitlement" would be vulnerable political game; but as a survivor of one boom and bust cycle, the basic idea makes good sense to me. Are there answers for individual scientists in these hard times? An insight from Mother Lode may help. It is the observation that scientists, especially physicists like me, are highly fearful that their research skills will become obsolete. (So strong is the reaction that the book pointedly iterates the obvious, that its analysis points out no particular person's obsolescence.) Reading this, I examined my own fears. As for my own obsolescence, save your condolences: Without eternally rattling the tin cup (but with the occasional discreet shake), I'm finding ways to do more of what I want, a reclaiming of my independence leading to new productivity. Other scientists can do this, too, by overcoming false fears of obsolescence engendered by our competitive ways. In fact, what are truly obsolete and need overhauling are those attitudes in the system that have ignored reality, most especially those of our hide-bound, turf-defending professional societies, and those found in the federal apparatus for science policy and funding (although this shows recent signs of life). Those should change, to determine what mix of kinds and amounts of science is truly needed, and to provide steady support for it. It is up to the individual scientists to turn the support into scientific achievement. This can come from funded research, yes, but also from dedicated teaching, enlightened administration, commitment to a true national science policy. The more we scientists find creative routes to career fulfillment through work of high quality, the more the professional societies, universities, corporations, and funding agencies must respond. With our own efforts, support from our profession, and a new administration in Washington, we can hope--along with the authors of Mother Lode--that future cohorts will see wiser support of science. Sidney Perkowitz is Charles Howard Candler Professor of Physics at Emory University, Atlanta. He recently spent a year at the Southeastern Universities Research Association in Washington, D.C., observing how science policy is made.

Generating Science : Productivity and Policy

Interested in reading more?

Become a Member of