Grant writing is doing science, but also often a waste of time
Scientists spend too much time writing grants and not doing science.
This is a common complaint, but it is a little simplistic. A lot of grant writing is science. In fact, I rarely do a deep dive into the literature anymore except when I’m writing a grant and this is very valuable for my science. When writing grants, I also plan experiments and data analyses, figure out what is important to work on and do many other important tasks towards better science.
The problem is that not all work towards a grant advances science as a whole. Some of it, really is just for this specific grant and is purely a cost of grant writing.
We can do a standard 2x2 matrix on useful for grant/science:
The green quadrant of things that are useful for both the grant and science is actually a large part of grant writing. The big problem is the red-colored top-left quadrant: work that is useful for the grant and only for the grant.
To a large extent, there are some funding best practices that minimize work in the red quadrant12:
Make formatting checks light and/or automated. People currently worry whether converting from Word to PDF in an Apple computer can result in text being submitted with a 10.98pt font size instead of the mandatory 11pt.3
Since most grants are going to be rejected, consider doing a two stage selection where an initial stage is much less laborious to apply for and details are only required for those selected for the second round.
A CV should be a PDF document, not a form. Particularly not one where you need to type in a bunch of irrelevant information: Is there a single instance where the exact date in which I was awarded my PhD was a piece of information that changed any single decision?
A much more controversial proposal though is to make grants more dependent on track-record. There is an ideal whereby grants are evaluated blindly: an idea should stand on its own independently of who proposes it: the same grant proposed by a Nobel Laureate and by a graduate student should have the same chance of succeeding. This is rarely the case, but I would argue that, if anything, track-record should count for more. Naturally, this needs to be adjusted for career stage: an applicant at the start of their career cannot be compared to someone with 20 years of experience and there is a risk that, done poorly, it leads worsening the “up or out” nature of academic careers that is so destructive.
However, done properly, it would mean moving as much as possible from the bottom-right (useful for science, but not for the grant) into the top-right (useful for both funding and science), while even further reducing the red quadrant.
Ideas are cheap and we all know that what ends up happening is often very different from what was written in the grant4: a grant written now cannot predict what will be the best science to be performed three or four years. In the extreme case, funding can be offered purely based on track record: here is what I’ve done in the past 5 years, can I pretty please have some science money?
Generally speaking, private funders tend to be better than public ones at this. If they have limits, it’s a textbox on a webpage that counts words, not someone with a ruler checking whether a PDF is conformant.
Another important question is: if the red quadrant is so important, how does this shape the culture of science but deciding both who is selected and what is rewarded?
This should be seen as an embarrassment by the science funders who impose and enforce such rules, but my guess is that they are proud of it.
Part of grantsmanship is writing in a way that gives you some room to do something different later if you realize that there are better opportunities.