#35 from R&D Innovator Volume 2, Number 5          May 1993

Age and the Quest for the Ribbon
by Paula E. Stephan, Ph.D., and Sharon G. Levin, Ph.D.

Dr. Stephan is professor of economics and senior associate, Policy Research Center, Georgia State University, and Dr. Levin is professor of economics at the University of Missouri, St. Louis.  They published Striking the Mother Lode in Science: The Importance of Age, Place, and Time (New York: Oxford University Press, 1992) from which the following piece was adapted.

Science is a young person's game: This is a view to which many scientists subscribe, although few take as extreme a position as the physicist P.A.M. Dirac, who said, only half-jokingly, that a physicist is "better dead" than living past 30. 

We have examined the relationship between age and scientific productivity.  Our quest was motivated in part by the sobering realization that the recent slower pace of science has caused the typical researcher of the 1990's to be significantly older than were researchers 15 to 20 years ago.  While our research focused on academic scientists, where success is measured by publications, the same conclusions generally apply to industrial scientists, where success is measured by patents and trade secrets, as well as publications.

We concentrated initially on explanations for an age-productivity relationship that might exist.  Many of the explanations we examined hinge on the factors that motivate scientists to do science.  These are best summarized as the puzzle, the ribbon and the gold.  A scientist hopes for intrinsic rewards from solving the puzzle, and for extrinsic rewards from recognition (the ribbon) and financial remuneration (the gold). 

Quest for the Ribbon

How can the quest for the ribbon cause a relationship between age and productivity?  By the time scientists take their first professional position, they have a fair idea of what they need for recognition:  publication in respected journals (or, in industry, recognition for successful innovation). 

In addition to effort, there are other determinants of success, particularly talent and luck.  Just as some musicians have a special gift that sets them head and shoulders above other musicians, some scientists have an innate ability.  Luck also enters the picture:  a serendipitous event in the lab, a referee or supervisor partial to a particular scientist's research, a competitor who hasn't had success.

At the outset of their careers, scientists believe they have a high probability of success.  For illustrative purposes, consider scientist Jones, who believes her probability of success is 70 per cent.  This is close enough to 100 percent to motivate her to work hard, but not close enough to make her think that success will come easily. 

Jones selects a question to study, organizes the research, and performs the experiments, but the experiment does not yield positive results.  She tries again, designing a new experiment.  This time Jones is successful and submits the results to a prestigious journal, only to find that a competitor had submitted a similar article two weeks earlier.  Things are not looking good for Jones. 

She changes focus, begins a new experiment, and spends six months in the lab, only to find that a basic assumption has been proved invalid.  Three research projects, no acceptances.  Surely it's time for her to reevaluate the probability of doing research that can be published in a top journal.  (Her industrial counterpart has been unable to solve various assigned projects.)

We could write a different scenario for Jones:  Her first experiment was successful and the resulting paper was accepted by a top journal, which led her to do additional work, which was also published in top journals.  By this time, with several significant pieces of research, she gains a small following, and is invited to conferences, where she learns who is working on what questions, which is a great help if she is doing research in a competitive area. 

Do I Detect a Trend?

By this stage of her career, colleagues are referring to Jones's work and deferring to her as an expert.  Now, when she writes another paper, even if a referee is doubtful about certain points, the referee may defer to Jones’s “expertise.”  And when she applies for research funds, the review board will be quicker to approve her project than if a comparable proposal has come from an “unknown.”  (The industrial researcher who has solved key problems is, in a similar fashion, more likely to be listened to by her colleagues and administrators.)

Clearly, the longer Jones works, the better idea she has of the probability of doing research deemed worthy of a top journal.  In the worst-case scenario, she begins to wonder whether 70 percent chance of success is a gross overestimation.  Not only does failure and rejection lead her to reevaluate the odds of success, it also makes her question whether winning is all that important.  As a result, she becomes less motivated to spend time doing research.  She finds, as her career unfolds, that the “game” of research is not what she expected while in college.

In the best-case scenario, Jones learns that she has a probability of achieving success that’s far higher than 70 percent, and this encourages her to spend even more time in the lab.  Why?  All those requests for preprints and invitations to conferences whet her appetite for even more recognition, until she becomes a connoisseur of recognition. Jones begins to dream of honors heretofore unimaginable. 

Success from Success

Sociologists of science refer to this self-reinforcing process as cumulative advantage, and argue that this process means that the recognition for a specific piece of research may depend on the reputation of the scientist.  From the point of view of journal referees and the broader scientific community, research that is really only average may—depending on the investigator’s reputation—appear more like an outstanding achievement.  

The implications for productivity implicit in cumulative advantage and reinforcement theory are clear.  For the individual, scientific productivity is consistent over time: Scientists who are productive early on remain productive later on, while those who are unproductive at an early date are likely to be less productive later on.  Some succeed for a while, it’s true, only to fail later.  But on the whole, success breeds success.   

Generally speaking, studies confirm this pattern:  Later productivity is heavily influenced by recognition of early work.  These results could be explained by the “sacred spark” hypothesis—that the population of scientists is heterogeneous:  Those with the spark are always productive, those without it are not.  No one would argue that talent is equally distributed among scientists.  But research gives credence to the cumulative advantage hypothesis without totally discrediting the sacred spark hypothesis.  The differential distribution of talent does not appear large enough to account for the vast inequality in research output that exists in science.

There are other reasons besides cumulative advantage for thinking that a negative relationship exists between age and productivity.  Suffice it to say they involve the quest for gold and the desire to solve the puzzle.  Another factor that comes into play is the relationship between age and creativity.  We leave that discussion for another time.

The main lesson here for R&D managers is this:  Let your staff feel that they are continually increasing their value.  Otherwise, they will get discouraged about their career potential and will have less enthusiasm for productive research.  You have to determine an appropriate level of responsibility for each individual, making it such that he or she believes the ribbon to be within reach.