ScienceHound has posted a new analysis related to the NIH budget and award policy. He's been beavering away with mathematical models lately that are generally going to be beyond my ability to understand. In a tweet however, he made it pretty clear.
As expanded in his blog post:
The largest difference between the curves occurs at the beginning of the doubling period (1998-2003) where the model predicts a large increase in the number of grants that was not observed. This is due to the fact that NIH initiated a number of larger non–RPG-based programs when substantial new funding was available rather than simply funding more RPGs (although they did this to some extent). For example, in 1998, NIH invested $17 million through the Specialized Center–Cooperative Agreements (U54) mechanism. This grew to $146 million in 1999, $188 million in 2000, $298 million in 2001, $336 million in 2002, and $396 million in 2003. Note that the change each year matters for the number of new and competing grants that can be made because, for a given year, it does not matter whether funds have been previously committed to RPGs or to other mechanisms.
This interval of time, in my view, is right around when the first of the GenXers were getting (or should have been getting) appointed Assistant Professor. Certainly, YHN was appointed in this interval.
Let us recall a couple of graphs. First, this one:
The red trace depicts success rates from 1962 to 2008 for R01 equivalents (R01, R23, R29, R37). Note that they are not broken down by experienced/new investigators status, nor are new applications distinguished from competing continuation applications. The blue line shows total number of applications reviewed...which may or may not be of interest to you. [update 7/12/12: I forgot to mention that the data in the 60s are listed as "estimated" success rates.]
Ok, Ok, Not much to see here, right? The 30% success rate was about the same in the doubling period as it was in the 80s. Now view this broken down by noobs and experienced investigators.
As we know from prior posts, career-stage differences matter a LOT. In the 80s when the overall success rate was 30%, you can see that newcomers were at about 20% and established investigators were enjoying at least a 17%age point advantage (I think these data also conflate competing continuation with new applications so there's another important factor buried in the "Experienced" trace.) Nevertheless, since the Experienced/New gap was similar from 1980 to 2006, we can probably assume it held true prior to that interval as well.
Again, first time applicants had about the same lack of success in the 80s as they did in the early stages of the doubling (ok, actually a few points higher in the 80s). About 20%. Things didn't go severely into the tanker for the noobs until the end of the doubling around 2004. But think of the career arc. A person who started in the 80s with their first grant jumped up to enjoy 30% success rates and a climbing trend. Someone who managed to land a five year R01 in 2000, conversely, faced steeply declining success rates just when they were ready to get their next grant 4-5 years later.
This is for Research Project Grants (R01, R03, R15, R21, R22, R23, R29, R33, R34, R35, R36, R37, R55, R56, RC1, P01, P42, PN1, U01, U19, UC1) and does not refer to the Centers or U54 that ScienceHound discussed. Putting his analysis and insider explanation (if you don't know, ScienceHound was NIGMS Director from 2003-2010) to work, we can assume that these RPG or R01-equiv success rates would have been much higher during the doubling, save for the choice of NIH not to devote the full largesse to RPGs.
So. Instead of restoring experienced investigator success to where it had been during the early 80s and instead of finally (finally) doing something about noob-investigator success rates that had resulted in handwringing since literally the start of the NIH (ok, the 60s anyway) the NIH decided to spend money on boondoggles.
The NIH decided to assign a disproportionate share of the doubling to the very best funded institutions and scientists using mechanisms that were mostly peer reviewed by....the best funded scientists from the best-funded institutions. One of the CSR rules, after all, is that apps for a given mechanism should be reviewed mostly by those who have obtained such a mechanism. You have to have an R01 to be in a regular R01-reviewing panel and P50/P60/P01 are reviewed mostly by those who have been funded by such mechanisms.
One way to look at this is that a lot of the doubling was sequestered from the riff-raff by design.
This is part of the reason that Gen X will never live up to its scientific potential. The full benefit of the doubling was never made available to us in a competitive manner. Large-mech projects under the elite, older generation kept us shadowed. Maybe a couple of us* shared in the Big-Mechanism wealth in minor form but we were by no means ready to make a play to lead them and get the full benefit. Meantime, our measly R01 applications were being beat up mercilessly by the established and compared unfavorably to Senior PI apps supported by their multi-R01 and BigMech labs.
The story is not over.
Given that I grew up as a scientist in this era, and given that like most of us I was pretty ignorant of longitudinal funding trends, etc, my perception was that a Big Mech was...expected. As in eventually, we were supposed to get to the point where not just the very tippy-top best of us, but basically anyone with maybe top-25% verve and energy could land a BigMech. Maybe a P01 Program Project, maybe a Center. The Late-Boomers felt it too. I saw several of the late Boomers get into this mode right as the badness struck. They were semi-outraged, let me tell you, when the nearly universal Program Officer response was "We're not funding P01s anymore. We suggest you don't submit one.".
AYFK? For people who were used to hearing POs say "We advise you to revise and resubmit" at the drop of a hat and who had never been told by a PO not to try (with a half decent idea) this was quite surprising. Especially when they looked at the lucky ducks who had put their Big Mechs together just a few years before....well there was a lot of screaming about bias and unfairness at first.
P01s are relatively easy for Program to shut down. As always, YMMV when it comes to NIH matters. But in general, I'd say that P01s tended to be a lot more fluid** than Centers (P50/P60). Once a Big Hitter group got a-hold of a Center award, they tended to stay funded. For decades. IME, anyway. or in my perception, more accurately.
Take a look at the history of Program Projects versus Centers in your field / favorite ICs, DearReader and report back, eh?
Don't get me wrong. There is much to like about Program Projects and Centers. Done right, they can be very good at shepherding the careers of transitioning / new scientists. But they are profoundly undemocratic and tend to consolidate NIH funding in the hands of the few elite of the IC in question. Often times they appear to be less productive than those of us not directly in them would calculate "should" happen for the the same expenditure on R01s. Such complaints are both right and wrong and often simultaneously when it comes to the same Center award. It is something that depends on your perspective and what you value and/or predict as outcome.
I can think of precisely one GenX Center Director in the stable of my favorite ICs at the moment. No doubt there are more because I don't do exhaustive review and I don't recognize every name to put to a face right off if I were to go RePORTERing. But still. I can rattle off tons of Boomer and pre-Boomer Center Directors.
It goes back to a point I made in a prior post. Gen X scientists were not just severely filtered. Even the ones that managed to transition to faculty appointments were delayed at every step. Funding came harder and at a delay. Real purchasing power was reduced. Publication expectations went up. We were not ready and able to take up the reins of larger efforts to anywhere near the same extent when we approached mid career. We could not rely upon clockwork schedules of grant renewal. We could not expect that a high percentage of our new proposals would be funded. We did not have as extensive a run of successful individual productivity on which to base a stretch for BigMech science.
And this comes back to a phenomenon ScienceHound identifies. The NIH decided*** to put a disproportionate share of the doubling monies into Centers rather than R01s for the struggling new PIs. This had a very long tail of lasting effects.
*I certainly did.
**Note: The P01 is considered an RPG with the R01s, etc, but Centers are not. There is some floofraw about these being "different pots of money" from an appropriation standpoint. They are not directly substitutable in immediate priority, the way I hear it.
***Any NIH insiders that start in on how Congress tied their hands can stop before starting. Appropriations language involved back and forth with NIH, believe me.