Venture Capital and NIH
National Institute of TechBros?
Note: This piece does not represent the views of Duke University, but instead are my views offered under academic freedom.
***************************
This is an interesting piece by Aishwarya Khanduja and Stuart Buck that argues venture capital provides some important lessons that could inform how the federal government funds research (they focus on NIH and life sciences). I believe that their piece is directionally correct, and given all the turmoil and upheaval of NIH in Trump 2.0 we might as well take the opportunity to look at how we could improve the funding of research and science in NIH, NSF and maybe DoD. If I were going to summarize some key ideas from their piece and add my brief thoughts it would be:
Reduce grant review times and normalize failure. For review times, from months, or with a revision or two of a proposal, years to weeks. This is needed. Normalizing failures would be a huge culture shift. I agree with the direction, and further think that NIH should routinize publication of falsifiable results, or inability to reject the null, etc. That work needs to be available if we are to learn from it. They note that venture firms expect 70+% of the companies they fund to fail, with some doing ok and maybe 1 in 100 being a great success. I think moving that direction (faster, more failure, don’t be so risk averse) is correct, but I am unsure of the correct breakdown and how it may differ across types of research.
They suggest considering funding of people/labs instead of proposals, which is how some venture funds decide to back given leaders regardless of the business area. I am not so sure about how this adapts, nor am I certain this is not what we already do to some extent via the scoring system for NIH for environment, which means a place like Duke almost always gets a 1 or maybe a 2 here (1 is best). I may misunderstand something about what they are asking for here.
Two areas that need explicit discussion. First, debate and a clear decision are needed on the degree to which competition between universities is crucial (versus specialization; UNC do thing A, F, K, Duke do C, L, T). Duke and UNC have a planned shared children’s hospital coming and so maybe there are opportunities for less competition and more focus in some areas? Such a change would need a safe harbor of some sort, as Duke and UNC settled a restraint of trade action from the Department of Justice that alleged the two colluded to hold down faculty wages (I got a $4,000 check!). Second, what role will intramural NIH research play moving forward. There are probably highly specialized areas of science that may need to remain done in this way, maybe even expanded. This issue is probably best linked to the one about competition v. cooperation in particular areas of research.
Finally, I confess do being a bit queasy of anything that sounds like ‘what NIH/research needs is more of a broligarch/techbro’ vibe. One fact to remember when we think about the brilliance of Silicon Valley and venture capital’s ability to achieve outsized returns is to remember the story of Silicon Valley Bank (SVB). The SVB did not fail because of risky loans and the like, but because they apparently did not understand the most basic aspect of running a bank—hedging interest rate and duration risk in their ‘safe’ holdings. It was banking 101 stuff that brought the brilliant guys down, made worse because the techbros had way too much money in one bank far above FDIC levels, and not enough normal deposits, and the rich guys, brilliant guys started talking (and tweeting) about it, and lo and behold they created a good ole fashioned 1930 bank run that was a fairly large shock to the U.S. financial system. Of course the FDIC stepped in, and the Federal Reserve opened a repo window.
Just a reminder that in the venture world, the profits are private, but the big losses are socialized. And NIH is in the business of producing public goods, that can and should be more efficiently moved into private ones. Let’s proceed with eyes wide open.
Books can be written about what is wrong with the NIH funding model that include an abundance of theatrics and pearl clutching about the excessive bureaucracy and the 100-pages of administrative information added to the 12-page science part, which always includes what we know to be a strong overstatement about how brilliant people get locked into their five year plan even when the first experiment that they do suggests that they should redirect (the redirecting happens all the time-while you can't say you are going to work on insulin resistance and then pivot to developing therapies for gliomas, NO ONE is going to complain of you discover something important about insulin resistance because you didn't do Aim 2 exactly as you said you would). And it is easy to look at specifics and say "right here, instead of doing X we should do Y" without looking at how you get to this point where your only choices are x and y. There is much to criticize here but you know for the last 30 years I have sat in meeting after meeting hearing about how the pay lines are going down, the average age at first R01 is increasing, R2s are now expecting their junior faculty to have NIH funding when they are up for tenure, renegotiation of indirect costs takes into account debt service on the new buildings that R1s are building and expanded research infrastructure so that they can aggressively hire tenure track faculty with ginormous startups on the assumption that these will be "paid back" in 5-7 years through new grants, expanding graduate programs that collectively keep graduating 10,000 PhDs in biomedical and biological sciences even though training funds (as opposed to research funds) have been pretty much flat since the numbers were more like 6,000, and even though some percentage of those graduates are not graduating into jobs that actually require a PhD and these large and growing programs (because the new faculty in the new buildings with the big startups "need" more grad students) "need" more administrators (and full disclosure-this has been great for my career) because the students expect a high level of service, and then they unionize because, you know, you are not meeting their very special service requests, driving up personnel costs and OK I will stop here. The short version of this rant is that universities are now hardwired systems of increasing fixed costs while appropriations stay flat or grow more slowly, and any change, no matter how rational, correct, or necessary is going to cause a good deal of disruption and collapse. And that is how my Tuesday is going.
Great takes! I agree that learning from VC/industry could help reform some of the processes of NIH. Your points at the end about the core function of the NIH should not be lost on readers. The NIH research infrastructure, both intramural and extramurally, serves the public. In many ways, they also serve the business community by supporting ideas outside of quarterly scrutiny.
One would think that having an environment where evaluation of progress can focus on things outside of pure ROI would allow people to think more creatively about their proposals. Creativity in the scope of the science is welcome. As you note, this also takes a reflection on the core functions of NIH.
Time will tell!