"Grant System Leads Cancer Researchers to Play It Safe" in today's NY Times highlights NIH's crazy risk averse grant making process.

...the fight against cancer is going slower than most had hoped, with only small changes in the death rate in the almost 40 years since it began. One major impediment, scientists agree, is the grant system itself. It has become a sort of jobs program, a way to keep research laboratories going year after year with the understanding that the focus will be on small projects unlikely to take significant steps toward curing cancer.

What seems strange to me is that the people in the article keep blaming scarce resources for the excess of caution. The NIH's budget doubled recently with no real change in the likelihood of funding for risky proposals - if anything, things are worse because of the choices NIH made for how to spend the money. And NIH now has $10B of stimulus money to spend in the next two years. How hard could it be to find some dollars for risky proposals in that pot? This is not about dollars - it's about a fundamentally risk averse culture at NIH.

NIH is trying:

The National Institutes of Health has started “pilot experiments” to see if there is a better way of getting financing for innovative projects, its acting director, Dr. Kington, said.

They include “pioneer awards,” begun in 2004 for “ideas that have the potential for high impact but may be too novel, span too diverse a range of disciplines or be at a stage too early to fare well in the traditional peer review process.” But only 3 percent to 5 percent of the applicants get funded. Now the institutes have decided to set aside up to $25 million for “transformative R01 grants,” described as “proposing exceptionally innovative, high risk, original and/or unconventional research with the potential to create or overturn fundamental paradigms.”

These look like attempts to shoehorn risky projects into the current system - the grants look relatively large, which has the dual downsides of attracting lots of applicants and making the reviewers cautious (nobody wants to be the one to waste $1M or more on a failed project).

I think NIH would have much more success getting transformative results with a tiered system: lots of small grants (e.g. $50K-$100K) to test ideas; a modest number of mid-range grants to further advance things that emerge successfully from stage 1; and finally, the more standard $1M-$2M R01s for things that succeed in the previous stages.

I think the key is having large numbers of small grants. This would provide a lot of freedom to explore, and when some things fail - and they will fail - you're not out that much money. The NIH's current strategy is like investing everything in bonds and blue-chips: treasuries, AT&T, and Wal-Mart. If you do that, you get a safe return, but there's a very real opportunity cost: you miss out on the occasional Google. Investing small amounts in lots and lots of risky ideas is like having a diversified portfolio of start-ups. Sure, there will be lots of Pets.coms, but you'll also get some EBays and Amazons that you wouldn't get otherwise.

I suspect that there are two main reasons the NIH doesn't do things this way:

• First, I think that the people at or near the top of the NIH food chain would see any shift in resources away from R01s as a threat. You'll hear cries of neglecting "serious" research or the like.
• Second, reviewing and managing proposals is a resource-intensive process. The NIH has a hard enough time with getting their existing grants reviewed, and increasing the number of proposals funded would make things worse. I think funding smaller proposals would be easier, since the bar for funding would be lower, but still, it's by no means free.

In any case, I think it's great that this is making the front page of the Times. I hope that somewhere in the Obama administration, someone is preparing to light a fire under the NIH to get on this.