Today was the second day of the AAAC meeting in DC; I participated remotely from NYC. The most interesting discussions of the day were around proposal pressure and causes for low proposal acceptance rates (around 15 percent for most proposal categories at NSF and NASA). The number of proposals to the agencies is rising while budgets are flat or worse (for individual-investigator grants at NSF they are declining because of increasing costs of facilities). The question is: What is leading to the increase in proposal submissions? It is substantial; something like a factor of three over the last 25 years. At the same time, the AAS membership has increased, but only by tens of percent.
With data in hand, we can rule out a few easy hypotheses: It is not people just sending in lots of bad proposals; proposals are still very good and now proposals rated Excellent and Very Good have less than 50-percent acceptance rates. It is not proposals by young people; the increase seems to be coming from senior people (people with tenure). It is not more proposals in any one year by the same PI; most PIs put in only one proposal to (say) the NSF AAG call in a year. What it might be is that people are not waiting three years between proposal submissions; I know I don't! 25 years ago, a successful research-active astronomer would put in one proposal every three years. Now I think most of us put in a proposal most years. That might account for it, but we don't yet know. It is not absolutely trivial to get the data.
In addition to getting data from the funding agencies, we also plan to do some kind of survey of the community. This will give less "hard" data, but will be able to give answers to questions that we can't ask of the data at the agencies, about things like motivation, perception, and reasoning among proposal PIs.
As an outsider several (certainly provocative from the point of view of an US-based researcher) options come to my mind that could somewhat alleviate the problem (not necessarily with satisfaction though).
ReplyDelete1. Do it like the ERC and impose a delay before submitting a new proposal. E: PI can submit the following cycle. VG: PI cannot submit for 1 cycle. G: PI cannot submit for 2 cycles. F: PI cannot submit for 3 cycles. P: PI cannot submit for 4 cycles. I bet that would easily half the number of proposals.
2. Put more money into consolidated multi-year grants that fund an entire department, like in the UK. That would drastically reduce the load put on individual researchers for writing grants. Which means more time for science.
3. Put a (low) cap on overheads. Say 25% for instance, like the ERC. Lower overheads means more proposals can be funded with a given budget. That would not make universities happy but that would force them to rethink what they do with this money and be more efficient.
Anonymous: I am more interested at this point in *understanding* the problem than *fixing* the problem. (By the way, most of the fixes you propose will not be considered by either NSF or NASA, for various reasons related to their missions.)
DeleteIt would be interesting to see if the distribution of funding is also changing. If senior PIs are putting in more proposals, and the senior vs junior success rate doesn't change (which I don't know), that would mean senior people were getting a larger share of the funding. Also, whether it's going to the same people: do a smaller number of people get N>1 grant? Basically, what's the Gini coefficient of NSF AAG funding?
ReplyDeleteGlad to see that someone is looking at the data. I have been skeptical that the NSF's call for people to only submit one proposal was going to do anything about the over-subscription - few people are masochists enough to submit many NSF proposals at the same time. I wonder how funding pressures have changes over the last 20 years? There is probably more data now floating around from archives as well as national and private telescopes that come with no money attached. Paying students and postdocs to work on that can drive N(proposals) up. Also, rumor has it that there are more people working in soft money positions now than there were 20 years ago, these money have to come from somewhere. Maybe data on the budgets of submitted proposals can give insights on what is it that we need more money for and why the numbers are up so much?
ReplyDelete