I am currently serving as a co-investigator on an R03 project. In NIH terms, this means a small, self-contained 2-year research project with an annual budget cap at $50,000 per year. As co-investigator, it provides me with 5% “Effort.” That is — this project is budgeted in such a way that I am expected to spend 5% of my time working on it. This works out to be 0.6 months per year, or roughly 2 weeks and 3 days, or 13 days. I was happy to help write my part of the project when the grant application was being submitted (“rising tides” and all), but I didn’t realize what I was getting into.
For this project, I am supposed to do sub-cellular fractionation followed by Western blotting on 3 regions from 50 mouse brains (25 per year). Each sub-cellular fractionation generates 5 samples (total protein, crude synaptic densities, large synaptic plasma membranes, pre-synaptic vesicles, endosomal vesicles). Given the limits of the ultracentrifuge and time it takes to process the samples, I can do 6 per day just to generate the samples. (This is an 8-10 hour day, too). So that is roughly 5 days to process one region of one of the cohorts — or 15 days to process all 3 regions from one of the cohorts. In all, this generates 375 samples. We can run 4 Western blots per week (roughly; it’s a 3-day process with lots of incubation times, if you try to do more, it is easy to mess things up); let’s say it takes me 2 workdays to do 4 Westerns (this is generous). At 12.5 samples per Western (12 on one, 13 on t’other — making all 25 from a cohort’s region/fraction on two blots), that is about 16 days (of more or less non-stop benchwork) to complete the cohort. Not to mention data analysis, optimization, instrument preparation, supplies management, the emails, the meetings, the organization, storage, labeling (very important), note-taking, and record keeping.
So, on paper, I was budgeted to do 31 days’ worth of work (minimum) with 13 days’ pay. This isn’t necessarily an expertise thing, either. It’s not like if I were better/faster, I could get it done sooner, these are the physical limitations of what can fit inside instruments and minimal incubation times. This is assuming perfect technique and skill is not a factor.
So what is going on? Why did we plan this so stupidly? The 5% effort was not arrived at because we thought it would really only take me a few weeks to do it. The 5% effort was back-calculated from the budget constraints and my salary. This is a really stupid way to do things. No company would budget their employees’ time this way, so why on god’s green earth are we in academic science budgeting our time this way?
Two reasons: 1) Money — Budget constraints, the NIH modular budget system, and we must fit a proposed project into the budget; 2) Peer review — proposing a realistic amount of work given a particular budget would not (and did not) pass peer review. Reviewers envision a particular amount of work for a project should be done in annual terms (and they are correct, in scientific terms).
The NIH modular budget was implemented in 2001. This sets R01 limits within a modular budget at $250,000 per year (5 times our piddly R03!). An R01 should have 3 specific aims, be roughly 3 times the scope of an R03, and go more in depth. In the old-school model, it would keep a team of roughly 5-7 researchers busy for 5 years. This is modeled around some organization of: One Principal Investigator, 1-2 post-docs, 1 lab tech, and 2-4 grad students and maybe an undergrad. In 2001, at $250,000 per year, this may have been feasible to support with an R01 plus some institutional funds (I and my cohort of Assistant Adjuncts n the UC System get zero institutional funds). This model, btw, has fallen apart with the adjunctification of the academy.
The modular budget limit of $250,000 hasn’t changed since 2001. But a lot has. Housing costs in major locations where biomedical research is being done have skyrocketed (Boston, Maryland, Southern California, Bay Area), and the cost of biomedical supplies has inflated as well. The NIH calculates a Biomedical Research & Development Price Index. Given the BRDPI, the graph at the top shows what $250,000 can purchase in todays’ biomedical research dollars; and what an R01 would need to be in order to have the same R&D effect as 2001. I am submitting an R01 application tomorrow, and I decided to exceed the modular budget because there’s just no way (especially considering the huge salaries of Senior Investigators) that really good, three-aim, five year, in-depth, sufficient breadth project can fit within this budget.
Another fun exercise is to see what %Effort this R03’s actual dollar amount would be getting in terms of dollars past. How far back in time do we need to go to get it to a reasonable number of effort to complete the project? That is — a minimum of 31 work days (1.6 months, or 13% Effort). I don’t blame my colleague for designing this project. It is worthwhile (if I didn’t believe that, I wouldn’t have helped), any money into the department helps. But the NIH, grant reviewers, program officers, and We Scientists need to give more serious consideration before we decide to design, budget, fund, and promote these projects and funding mechanisms.
I, myself, was principal investigator on an R03, but it was a special funding program with $150,000 per year, so I was able to do a lot more with it. However, I have concluded that these small grants are not worth it. They aren’t worth it for us to slave over because the purchasing power isn’t enough to obtain resources to do worthwhile research that is publishable in top-tier journals and get us tenure. It’s essentially getting our CVs at Costco. It’s not worth it to the NIH and the greater mission of academic biomedical research for the US either — it produces dis-jointed projects and careers that lack scientific continuity.
If the stagnant NIH budget isn’t going to rise, I see several solutions to these problems: 1) Fewer PhD students, 2) Retirement at 65, 3) Get rid of these stupid small grants, 4) Universities must be required to commit salary to principal investigators, 5) Grant reviewers need to be realistic.