Ratio of <= 40 to >= 50 scientists funded by the NIH

The dotted line at the top is the Jones-implied ratio of productivity of <= 40 year olds to >= 50 year-olds, as drawn from Figure 1 in this source.

For the construction of this data source I am indebted to PseudoMontaigne.  Does it not imply that NIH funding is vastly over-allocated according to the criterion of seniority?  Or might this be the rise of the lab system, where the older people are the PIs, and they in turn dole this money out to younger researchers?  More middlemen, so to speak.  Opinions?


Why not look at the external funding of various members of the GMU econ dept. by age, and see how it compares? After all, you are already intimately aware of major sources of such non-Commonwealth of Virginia funding, whether it involves the GMU Foundation (as noted in court documents) or of various public policy institutes that are legally separate from GMU.

Dude, why do you do this to yourself?

Why not? I get up, drink some coffee while listening to wfmu.org - on my second cup, after spending the last half hour doing various things like eating breakfast, lighting the woodstove, reading various other web sites - and comment as I wish. Admittedly, I am a disloyal reader, though strangely enough, 'loyal reader' seems to have fallen out of favor, possibly because it does not provide the necessary contrarian gloss. And Prof. Cowen is truly in an excellent position (quite possibly better than whoever is currently chair of the department) to know about such funding.

Of course pointing such things out won't change anything, but so what?

We all have different ideas of what is amusing, and this is one of the most amusing places on the Internet. In large due to nostalgia, actually - the GMU shell game as played among those that have now mainly settled into place at the Arlington Campus has remained unchanged over decades.

Sure, guy.

From my anecdotal evidence (I am a bio-tech scientist in Russia), yes, we have a lot of middlemen. Senior scientists, who produce general politics and ideas of a lab, while doing almost no hands-on work, and usually spend most of their time fostering relations between different labs and different sources of money. Do not know, if that's bad (seems to work atm) or if that anecdotal evidence is true in more general terms.

Interesting comment, thanks.

My experience is from an NSF- and private foundation-funded lab, but this rings true, and I observed that this was *not* because my PI *liked* spending all his time hustling for funding. The process selects out researchers who *don’t* do this. I don’t have a mental model to explain it.

Different setting, similar outcomes.

There is probably great value when the more senior, though perhaps less current knowledge, scientists serve such a role. The flip side is the rent-seek, agenda setting side where that position's power is abused for personal rather than scientific advancement.

I don't think one gets anything from the statistics offered to shed light on the good-bad evaluation of the situation. Perhaps merely a pointer saying maybe interesting to look into.

I am skeptical that the ratio is as low as indicated on that graph. (The actual situation is bad, but it's not *that* bad.) For better graphs, see https://nexus.od.nih.gov/all/2017/03/06/mid-career-investigators-shifting-demographics-nih-grant-recipients/ . I'd be curious to know PseudoMontaigne's data source.

Thanks, those are indeed better graphs. But only back to 1998, where you can't really see anything on this graph's scale.

They have been there their whole life, institutionalized.

I moved into consulting after the PhD . I enjoyed the research experience because my professor managed all the money issues and we could just focus on the job.

Still have friends doing research and they miss the time when someone else did the "get funding" work. Scientists like to do science not management or apply to funds. That's why motivated researchers love middlemen such as professors, institute leaders, department managers.

So, are they evil middlemen or simply an outcome of "division of labor"? Management is boring, but someone has to do it.

I understand the y-axis to be "# w/ funding 40" and that this isn't a proportion "% of researchers 40 w/ funding).

Because of that, we can't rule out that there are just differences in the underlying number of 40 researchers. It could be that their likelihood of getting an award is similar, but there are far more >40 researchers applying for grants. This could emerge if there is a trend away from NIH funded fields for prospective researchers. Or if new researchers changed careers more frequently to non-research sectors.

On a similar note, it could also be that the number of awards has changed. This could notably impact (or be impacted by) the type of awards offered. Since all researchers are not all eligible for all award types, this could also influence this ratio. If for example there was a trend away from entry-level grants (K-awards, R21, etc.) towards larger grants (R01) then something like this might emerge.

Yet another explanation is that at least some of the cohort of 40 cohort in later decades. They may well have been excellent grant-writers, and now have a history of successfully receiving grants - increasing the competition for new <40 grantee hopefuls. This could also be related to the point above regarding retention in the field.

There's probably a laundry list of other related explanations, but I think the graphic would be more compelling if it represented first a proportion of the underlying eligible population. Further, I think it could be supplemented by looking at award amounts, or $/researcher. As I presume that younger researchers may have even have more awards which are smaller in amount (R21, K awards, etc.) while older researchers have fewer, but larger awards (R01).

Is capital requirements lower today than 40 years ago?

How much did a high quality microscope cost 40 years ago compared to an xRay scanning system or a gene sequencer?

Do NIH grants include million dollar plus capital budgets for acquiring capital if the researcher doesnt have capital, or is the researcher with 30-40 years experience building up a lab, perhaps even designing, and then contracting to have his lab equipment sold to lots of privately funded researchers, getting equipment "free", as part of testing prototypes or first run equipment.

Half a century ago, labs had associated equipment production, whether glass blowers or metal fabricators and lots of skills to construct control systems, plus sensor and recording equipment. To get a sense of what most labs had to be able to, get a copy of the complete SciAm Amateur Scientist columns which provided cookbooks for building equipment for experiments based on past research equipment production. At the time of each column, the equipment could be ordered for prices far in excess of the price an amateur could afford.

Even today, you can do genetic work manually at home,taking days to do what you can dial into hundred thousand to million dollar machines and complete with less than a second of human effort, so research can do millions of experiments when in the past it would be hundreds.

The accomplished 30 year researcher gets a position that includes building a lab, but he must win over rich guys who build the building, or clean room lab, they get to name, while recruiting students and mentoring them, along with the funding to support them.

Research today is like automaking today, not the state a century ago where bicycle makers and horse carriage makers were building motor cars powered by DC electric motors, steam power and alcohol burning IC engines.

Cowen has many blog posts commenting on the slow-down in new discoveries - a Google search will reveal dozens of articles on the issue. Maybe it takes longer today, so funding from NIH comes later in one's career. Combine that with the middle-man phenomenon, and life begins at 40 for the scientist. I'm a lawyer, so it's not science, but law review articles are often researched and written by students but credited to faculty. Court opinions are more often than not researched and written by clerks even though credited to the judge. In private practice, the actual work, including the due diligence and drafting contracts, in complex commercial transactions is performed by junior lawyers in the firm but it's the partner who gets the (fee) credits. This arrangement (young lawyers do the work and senior lawyers get the (fee) credits) was the almost universal arrangement for law firms, including the large firm where I started my career in private practice. It worked because each generation of lawyers benefited from it: as the young lawyers aged, they assumed the position of senior lawyer and got the (fee) credits. This arrangement has been displaced in many if not most firms, because in the 1980s and 1990s firms began doing the unspeakable: poaching other firms for "talent" (i.e., lawyers with fee-paying clients). What became of the senior lawyers at such firms? They were sent to pasture. For the lawyer, life may begin at 40 but it ends at 55 or 60.

The internal contradictions of the American regime are becoming more and more obvious as the time goes by.

As we all know, sexual harassment in the workplace has become a hot topic (Cowen blogged on the #MeToo movement this past weekend). I am a lawyer, and we are subject to a code of ethics, violations of which can result in disbarment. Well, most states' codes of ethics for lawyers prohibit the lawyer from having sex with a client. [Some codes go into great detail about the meaning of "sex" while other states leave it to the imagination.] My reporting period for CLE is coming up, so I spent most of this past weekend viewing podcasts of CLE lectures. My state requires a minimum number of hours on ethics, so several of the podcasts were devoted to ethics. A two-hour ethics lecture I viewed this weekend was devoted exclusively to the ins and outs of when a lawyer can have sex with a client. It was presented in a fashion similar to a CLE lecture on how a client can avoid tax by deflecting income to a tax haven. The lecturer even included a libertarian angle by alluding to "freedom": the client's freedom to have sex with his/her lawyer. Somehow I don't believe this complies with the spirit of the #MeToo movement, but I wouldn't know about such things.

Honestly, it's a joke that people "can't" have sex with their lawyers. The way all lawyer rules are written is based on the assumption that lawyers are godly and clients are worms, and we have to protect the worms from being taken advantage of by the godly lawyers who would otherwise have their way with them. Far from the truth, in my experience.

Can I just jump in here on an unrelated topic. Several months ago, Tyler was a promoter of this idea that the Cubans were using sonic weapons against the US embassy in Havana. Now it turns out that this may have just been crickets. Is Tyler just a big fake news poster? Will he re-evaluate this?

Two possible contributing factors:

1 - There was a boom that is now playing out where more new scientists were younger. Now the younger scientists are more likely to leave academia.

2 - The length of time to become a professor increased. (Posrdocs werent as mandatory)

'Will he re-evaluate this?'

Wouldn't the standard Internet answer to such a query *crickets*?

My impression is that in 1965 rock stars were about age 24 on average but by this century it was assumed that could take well into your 30s to become a rock star.

Agent: It seems unlikely to me that there were "Indies short-tailed crickets" in all the places where people reported the symptoms. Further, if crickets can cause these symptoms, why isn't it well-known? People all over the crickets' range would have been reporting it to their doctors, not just employees of the US Embassy.

Since the situation was localized we are left with the possibility that if it was cricket sounds, someone was deliberately playing them with the goal of driving Embassy employees nuts. Instead, many articles (and your post) imply that it was something that "just happened," rather than that it was a planned attack.

Suppose a handful of people managed to corner the market 40 years ago, when they were young, and then they just kept their positions of power for the next four decades. The graph wouldn't look much different under that scenario, would it?

My sample size: 2. Both medical research labs, one run by brother-in-law, one by a neighbor. Both persons have told me that, as number of science PhDs grows faster than NIH (etc.) funding, competition for funds increases. Writing grant proposals is a skill like any other, you get better at it over time, and labs have to, given such increased competition. Better grant writers ( = scientists who pull in more money ) are steadily promoted in labs, until best writer ends up as head of the lab. This takes time. As a result, all the lab's grant papers are at least co-written by the senior person. Does this explain the chart? I don't know, I will write a grant proposal for a project to find out.

Winner for the day! (Exactly what I see also.)

Isn't it a sign of the general tendency to complacency and corruption as human institutions age? Especially given the transformation of science over several generations from an obsession of a few to a major mass-production industry.

If I were given to fretting over Science, I'd worry about theoretical physics being stuck since the 70s, the golden age of medicine being long behind us, and the low truth-value of such mass campaigns as statins-for-all, eat-no-fat, and, of course, Catastrophic Anthropogenic Global Warming.

Against which one could balance the triumphs of genetics, while chortling at the efforts of geneticists to push their discoveries while denying any possible un-PC interpretations of them.

Illustrates predatory Boomer management and administration skills, someone might say.

If the graph charts grant-recipient ages, what data chart the relative ages of the grant-makers?

Disclosure: NIH funded researcher; member of an NIH study section; age > 50.

When a study section reviews a grant, the first consideration is the scientific importance and rigor of the study design. But another important consideration is whether the investigative team is qualified to pull it off. After all, if it's innovative research, it means that nobody quite knows how to do it exactly, so you have to see if these people look like they can do what they propose.

Now, a few decades back, a typical R01 grant proposal allowed 25 pages to explain your proposal. This left plenty of room to provide a great deal of detail, and to say explicitly how you might handle certain foreseeable obstacles should they arise. But the current rules reduce this to 12 pages. So really all you can provide is something of an outline. If the resumes of the investigators show a lot of experience and successfully completed projects, the reviewers will tend to assume that they can handle things. If the investigators are relative novices, then it is something of a crap shoot whether they can do it--and the proposal itself gives them no opportunity to explain why or how they can. This clearly gives older, experienced investigators an edge.

In fact, for younger investigators submitting proposals that are on or near the cutting edge, I think the best thing they can do to boost their chances is to recruit a prominent researcher as a co-investigator, and preferably one with whom they have collaborated previously.

Obviously this is only one piece of the puzzle, but I do think the shortening of the proposals has forced reviewers to rely more on the experience of the team. Others have pointed up other contributing factors.

Capture of rents by established players?

10 points for you -- you said the same thing I did, only you did it much more concisely and eloquently.

Must be a slow news day. NIH is well aware of this - and has been for some time. In fact they've proposed various schemes to address the problem. (None of which are without serious side-effects - at least on paper.) Gov't Funding is arguably a zero-sum game. There are undoubtedly too many PhDs in Medicine "related" fields (biophysics, bioelectronics, biocomputation, etc. etc.) relative to the available funding, and I'd guess the ratio of active PhDs to available funding has gone up as the fraction of GDP spent (wasted) on human health & medicine has. They've had to limit the scope of use for grants since the parent institutions have been increasingly using Hollywood accounting to charge researchers for "overhead". There's been a lot of research done on this as well. One suggestion has been that applications should be anonymous because of the bias (i.e. old-boy (literally) system), this idea has persisted even though it is obvious that the experts sufficiently skilled in the art to vet the grant proposal will usually recognize which lab it's coming from simply by the details contained in the document. There's still a lack of consensus about whether younger researchers are actually more creative. It isn't as obvious that 25 years of education is sufficient for most PhD level students to master their trade. Science is one of those areas of knowledge which builds on what has gone before in such a way that requires the student to be conversant in much of the material. (What would automotive engineering look like if engineers were required to understand - in detail- steam power and buggy whips? (although this overstates the difficulty somewhat). I've also seen age distributions of the teams evaluating the proposals; and guess what? The most promising proposal (imho) has been a sort of crowd funding, where the 'pool' of proposers rank their peers proposals (the obvious problem is tit-for-tat voting). The idea that oldsters have too much power in the current system is appealing, but I suspect that the current system is just too damn big to be effectively 'managed' outside of institutional processes (which inevitably leads to accumulations of power and influence). The other problem is the goal of the funding: is it to make a marginal addition (accumulation) or to revolutionize? Is a grant leading to 5 publications more 'valuable' than one which leads to zero? How many patent examiners would be awarded a grant to study topological manifolds?

Age of recipient of grant is not a measure of anything, other than input. Look at output.

Maybe look at patent grants and papers resulting from grant, and the mean age of authors.

The other thing you might want to look at is that the grant process is a network process...people who get grants know those who give or review them, so it is not surprising that there would be a tilt toward age.

The other thing you might want to look at is the y axis which is ratio of late year to early year and the change in mid 80's. Is this the result of a decline in research dollars to fund post-graduate work. In other words, was the previous period one in which funding post-grads through grants more common and used as a way to subsidize education.

'Iron Law of Oligarchy'

... what was James Madison's 1789 vision for the NIH and Federal Health Care agencies?

In STEM, it comes down to lab structure; most labs have a PI, typically at an advanced stage of their career, whose basic function is to write grant proposals. They're competing for grants with all the other PIs of other labs, and all of these PIs are the best-and-brightest and most go-getter individuals in their fields, so this competition is brutal and time-consuming. Also; grant-writing, though everyone hates it, is a skill, and one which these PIs have perfected to an absurd degree; a newly-minded PhD independently submitting a grant application into this pool is typically like a bright-eyed karate student going up against a bracket of wizened old judo masters.

So it's no surprise that the age of the grant "recipient" is going up, but at the same time, the actual *work* that that grant funds (and the paper(s) that come out of it) are going to be largely done by a postdoc or grad student; in terms of individuals performing the hands-on scientific labor that the NIH is paying for, age is likely trending younger over time.

This situation is an equilibrium that everyone dislikes, but it does have some advantages- namely, it at least partially insulates everyone who isn't a PI from the funding process, so they have some time to actually do science.

A ratio (your y-axis) should almost always be plotted on a log-scale. As shown, I can't really tell what the values are when they are below 1.

Not in the US system but do work in Australia. What has tended to happen in my fields - Immunology, stem cells, dev bio etc. is that you tend to get a brilliant scientist as a lab head with all the ideas. You get a bunch of technicians, research assistants and post-docs to carry out the vision. If they are successful and publish well the lab head will often hire a post-doc who is good at writing grants to get more grants whilst they might target other sources of money or develop collaborations etc. After a while some of the post-docs in their mid to late 30s come back from the US or Eurpoe, have good ideas of their own and these days tend to get early career grants and make a sub-lab within the larger successful lab.

Eventually, the initial lab head gets older and one of a few things will happen. They either retire, commercialise their own work, might get recruited by a pharma company or another University, or move into a different role at the University. At that point the sub-labs make become full fledged labs.

Comments for this post are closed