Via Tyler Cowen, comes this graph of demographic shifts in NIH grants, which show a clear trend: older scientists are getting more money.
Cowen also cites the eminent economist Paul Romer, who worries about the effect of this shift on innovation:
Instead of young scientists getting grant funding to go off and do whatever they want in their twenties, they're working in a lab where somebody in his forties or fifties is the principal investigator in charge of the grant.Â They're working as apprentices, almost, under the senior person.Â If we're not careful, we could let our institutions, things like tenure and hierarchical structures and peer review, slowly morph over time so that old guys control more and more of what's going on and the young people have a harder and harder time doing something really different, and that would be would be a bad thing for these processes of growth and change.Â I'd like to see us keep thinking about how we could tweak our institutions to give power and control and opportunity to young people.Â
The bad news is that Romer is right, at least in part: young scientists, in general, tend to be a bit more innovative. (If you noticed all the conditionals and hedges in that sentence, please keep on reading.) The ingenuity of youth is perhaps best demonstrated by the inverted U curve of creative output, a well-studied phenomenon in which creativity rapidly increases at the start of a career before it crests and declines. Here's Dean Simonton, a psychologist at UC-Davis who has painstakingly quantified this demographic data:
One empirical generalization appears to be fairly secure: If one plots creative output as a function of age, productivity tends to rise fairly rapidly to a definite peak and thereafter decline gradually until output is about half the rate at the peak.
For instance, Simonton and others have shown that physicists tend to make their most important discovery before the age of 30, which is why they morbidly joke that if they haven't done Nobel-worthy work before they get married, they might as well quit the field. (The only field that peaks before physics is poetry, with an ideal creative age of 21.) Simonton argues that young physicists and poets mostly benefit from their innocence, ignorance and naivete. Because they haven't become "encultured," weighted down with false assumptions and tedious obligations, they're more willing to rebel against the status-quo. (Simonton rejects the obvious alternative explanation, which is that the creative decline is due to age-related cognitive decline. After all, some academic fields, such as literary criticism, have a peak creative age in the late forties.)
So I do think Romer is right to worry about this slow creep in grant funding to older scientists. However, before we start blaming the staid conservatism of the NIH, I think it's worth considering the extent to which this shift might be due to intellectual changes within scientific fields. In other words, the changes in funding might be a side-effect of scientific progress, and not an institutional failure.
Let's begin by considering the differences in peak creative age between different scientific fields. While physics, math and poetry are dominated by brash youth, many other fields are more amenable to middle age. (Simonton's list includes domains such as "novel writing, history, philosophy, medicine".) He argues that these fields show a very different creative curve, with a "a leisurely rise giving way a comparatively late peak, in the late 40s or even 50s chronologically, with a minimal if not largely absent drop-off afterward." (These differences are also cross-cultural: for instance, the age gap between the creative peaks for poets and novelists has been found in every major literary tradition across the world, with novelists getting wise and poets getting stale.) This suggests that the most efficient allocation of grants in these fields - at least if we want to fund innovation - is to fund medical researchers, philosophers and novelists in middle age, when they're tenured and deeply "encultured". Sometimes, innovation requires decades of education. That might not be romantic - it's amazing how many cliches of creativity come from 19th century British poets - but it's the demographic reality.
What accounts for these stark differences in peak creative age? One possibility is that they are caused by intrinsic features of the academic disciplines. As Simonton notes, those disciplines with an "intricate, highly articulated body of domain knowledge," such as physics, chess and poetry, tend to encourage youthful productivity. In contrast, fields that are more loosely defined, in which the basic concepts are ambiguous and unclear - examples include biology, philosophy and lit crit - lead to later peak productive ages. Furthermore, the peak of all intellectuals seems to be getting postponed, as the increasing complexity of research in general requires increased time to master. In 1500, the peak of creative output was 25; by 1960, it was 37.
But that doesn't mean we can afford to ignore the graying of NIH funding. I've talked to far too many young scientists who are exhausted by the bureaucracy of getting money. We need to fund impetuous youth, if only to give them time to grow old and wise.
- Log in to post comments
Do we know whether the age distribution of the applicants has changed?
Very nice post, holmes. The reason practitioners in fields like biology have their creative peaks later in life is because that creativity depends much more on the accumulation of breadth and depth of knowledge and the ability to make unexpected connections between different things, while in fields like mathematics and physics it is much more about sheer intellectual processing power. Poets peak early because poetry is mostly about being horny!
Thank you, Rosie Redfield. That's a big honking question that needs to be answered before we make too much of this graph. How much of this age shift is just the baby boomers coming through, with the baby bust in their wake?
The other obvious trend, which neither of you mentions, is that the distribution is become flatter and better rounded. I need to think about it some more, but that particular trend seems like a good thing. Although that too could just be a result of the current Boomer Echo wave having a wider spread than the one before it. That peak in the 1980 numbers is just crazy! Talk about turning 40 being the professional kiss of death.
Also, with regards to possible demographic shift over time, note that the peak is creeping forward -- but trailing substantially behind the actual passing of time by about two-thirds. So if the Boomers really are significantly skewing the demographics of the applicants, which seems likely, then correcting for that influence, any age-bias in the grant process itself might actually be in the opposite direction over the longer term. Yeah, I'm speculating wildly on this point; again, we'd need to see the data.
Or maybe 40 is just the new 30?
The Boomers. Definitely a problem. But it begs the question of how and why this occurs... If the implications for innovation advanced by JL are really that dire, structural solutions are warranted.
Simply not TRUE. Mathematics, which arguably requires the deepest, most complex understanding to produce notoriously peaks Extremely young. With the older researchers often joke about turning to performance enhancing drugs to keep up. Furthermore, data on 'peak creativity' are themselves biased by the funding stream.
To begin with, A good article and meaningful article. While I agree that we must fund our young scientists, I have a bone to pick with one of the assertions made herein. I cannot, and will not believe that a poet produces his best work at or around the age of 21. Being a former poetry MFA student I have read and written a large amount of poetry, and I have learned something about the education and ability of poets. A poet can be good at 21, even amazing, but they cannot maintain that throughout an entire collection. Like anything poetry requires hard work, revision, but mostly knowing and embodying the literary work that has come before you. I don't see these things coming out of our 21 year old poets. Though the Rimbaud's of the world exist, the great poets of the 20th century have been more like Yeats, John Berryman, or Wallace Stevens, who produced their greatest work well after they left college.
NIH funding is down 50% (20% funded, vs 30% funded) from not that long ago. What I think you're seeing is that OLDER scientists write better grants due to their developed skills versus those undeveloped skills in this process of application and funding.
Also, it seems that things have changed quite a bit in academia. Many professors are continuing with active research longer. When my wife got her PhD, you were pretty much expected to be teaching by 40. Now, at 50, she and many others are still doing science and not teaching.
Not so much because they want to keep researching, though many do. But because the way States, mostly Republican run States, have butchered academia.
I know my program (in accounting) was a top-10 program. But then a wave of Republican governors have destroyed its funding and it's...
Well, no kind way to put it, it's flat-out mediocre. The professors that made it a great program either retired, fled to industry or were recruited elsewhere. The graduate students flock to other programs. The top students no longer register on the national scale and are no longer recruited by other programs.
It's become a complete joke. Just another crappy California State University staffed by second-rate professors teaching in Cattle-Call Classes.
I don't like this trend, but to some extent there are practical reasons why it might be expected and welcomed. One of them (for the field I work in, neuroscience) is that, in general, the questions scientists are attempting to ask today are significantly more complicated and require more depth and breadth of investigation as compared to 30 years ago. This means that pursuing answers to today's scientific questions requires a large number and broader range of highly specialized techniques and expensive equipment.
Senior researchers don't always know how to perform more types of experiments than greener ones, rather, they have the money to recruit and pay multiple post-docs who DO have the skill sets OR the connections/clout to get other laboratories to do the far-flung experiments for them. Basically, the average young investigator has a hard time getting a foot-hold in the big money arenas (NIH-R01 grants) because he or she cannot publish in the big journals, because to publish there the research must be multifaceted and cutting-edge. Young investigators do not have the man-power (post docs), the equipment (usually acquired over decades in a multi-renewed R01 lab), the skill-sets (measured in # of post-docs and connections with other laboratories) OR the time (new faculty members at academic institutions are often loaded with administrative and teaching duties to ease the loads of older faculty members who have "paid their dues") to power a pair of high-level publications into the limelight just in time for their first grant review.
Basically, the reasons for this trend seem practical. But, just because a system trends in a particular direction, doesn't mean that it is all okay. In any given field, novel ideas coming from the younger generations always seem to be of an indefinably particular sort; not always, but usually more (if not awkwardly) progressive. If the graph keeps shifting rightward, what bright young college scientist is going to be excited about a field in which the odds of financial support and tangible success are stacked against him/her?
How has the changing role of women in research (and family) affected this chart and the assumptions made in JL's article? In the past 30 years women have become more proportionately represented in some fields than others. In addition women are more likely to return to work sooner after childbirth than in the past, enabling them to continue to pursue serious research after having children rather than dropping out altogether. Use of birth control has also become more reliable and more socially acceptable both in and outside marriage allowing women to delay having children until later in their careers. I don't have access to any data, and these factors may well cancel themselves out, but I have to think it's worth looking into as a potential driver, as well.
I'd quibble with the argument that poetry is a young person's game. What is young? By modern standards Shakespeare died at young middle age, but in his day, he was a modestly old man. And his best (mature) stuff came late in life. No doubt, many poets have burnt out young, but who's to say that Dylan Thomas' best days were not before him?
Moreover, the poetry market is poorly defined. What is the canon of great ones? And how is greatness determined? Physics seems a different matter, but still problematic to measure. What constitutes greatness, or even quality? Publications in peer reviewed literature? Nobel's? Grants? Patents?
It's also possible that there have been fewer hires in the past decade or two so most professors are older. Several other comments suggested broader demographic trends, but reduced funding for academe and the shift to adjunct labor seems a likely explanation for many of the changes in distribution of these grants.
It makes sense to me that science is a younger person's game. Between running the experiments, trouble shooting the results and running the experiments again, there's a lot of time and energy needed. Folks under the age of 35 typically have fewer commitments and energy to burn so they'll have a better chance of striking scientific gold.
Also, after doing science professionally for 15 years I think most scientists would find themselves in a comfortable routine that typically involves less profound tweaks on their earlier, more fundamental discoveries.
The same idea explains poets' early peak as well. After writing poetry for a few years, I'd think most writers would be hard pressed to a new groove and typically produce works that simply echo their earlier prose.