The following email conversation involves messages between Eli Kintisch, the editor of this blog, and scientists Christina Smolke and Natalie Mahowald. The issue: the enormous pressures facing scientists in the 1st decade of their careers, when they must balance writing grants, starting up a lab, and publishing key papers to get established in their careers. For more on career issues see ScienceCareers.
Dear Christina and Natalie:
It is a summer day in 2009 in Cambridge, England, and K. (39) looks out of his lab window, wondering why he chose the life of a scientist. ...The second half of 2007 and all of 2008 had been a nightmare—14 of these 18 months had been almost entirely devoted to writing grant applications. K. now sees how he has changed from being an enthusiastic scientist into an insecure bureaucrat. He feels he has lost much of his last 3 years and wasted his [British government] grant, despite doing his very best. … the present funding system eats its own seed corn.
That’s a bit from the sordid story published today in PLoS Biology of the anonymous "K.", a British colleague of Peter Lawrence, a senior scientist at the University of Cambridge. Lawrence, whose PLoS piece describes K's plight, blames the "black arts" of grant writing required by government bureaucracies on either side of the Atlantic. He bemoans a “Kafkaesque” regime which he says requires too much "salesmanship and networking," application forms that are too long, and as a result dooms too many smart beginning scientists' careers.
Thanks to both of you for taking some time to share your thoughts on your experience being a young scientist and balancing the care and feeding of your lab's bank account with, well, doing science. Christina, you're a biochemical engineer at Stanford University in your 6th year in a faculty job; Natalie, after a few years at the National Center for Atmospheric Research in Boulder, Colorado, you received a faculty job at Cornell University last year. I'm eager to know: do you find it difficult to find time to actually mull ideas? Do you write more grant applications than papers?
And what do you make of Lawrence's ideas on how to fix the system? He calls for shorter grant applications, grants which last longer, with easier renewal process. Or should young scientists simply be given bigger stipends and rewarded with funding based on publishing performance? NIH has lowered the length of its grant application from 25 to 12 pages, and some funders, like Howard Hughes medical Institute, have set strict limits on the number of papers that one can submit with an application. Is that enough Christina? How does it differ for the physical sciences, Natalie?
"I do often times feel like I don’t have time to sit down and think"
Dear Natalie, Eli:
My own experience in funding my independent research laboratory has not been as extreme as that described in this paper. I run a laboratory of approximately 15 full-time researchers. I do spend a lot of my time writing grants and fundraising for my laboratory. To support a lab this large I have found that it is important to be able to apply for and access funding from many different sources. For example, universities often provide different opportunities for internal funding, including funding for researchers in the form of departmental support for graduate students, training grants, and university and departmental fellowships for graduate students and postdoctoral researchers. I have found that it is important to be able to fundraise from many different external sources, including government agencies, private foundations, and private companies. One way to deal with the low funding percentages (and resistance of certain sources to fund more than one grant from the same investigator), is applying to many different sources.
One of the points I did resonate with in this paper was the frustration with the time period associated with when grants are submitted to when you are notified of whether or not they will be awarded, relative to the length of period of a standard award. This introduces a lot of uncertainty into the system, such that many investigators feel like they need to be continually writing grants, anticipating having to resubmit failed grants, in order to not get into a situation where there labs are low on funding and cannot support hired researchers. I do feel that I spend a lot of my time writing grants. I do often times feel like I don’t have time to sit down and think, but I believe this is due to all of the cumulative responsibilities that are asked of us as professors.
In addition to writing enough grants to fund a research laboratory, we are also asked to teach courses for our department, serve on committees for our department/school/university, review grants and research papers, write research papers and reviews, travel to share our research in different venues, and many other ad hoc responsibilities.
I agree with Lawrence’s suggestion on grant award periods - making grant periods (and renewals) a little longer (5 years) would be a big improvement to the current standard of 3-4 years. I don’t necessarily agree that shorter grant applications is the way to go. I think if grant applications are too short (~1-5 pages), one is often awarding a grant based on the track record (or reputation) of the investigator or the general idea that is being presented - as this is often the format that is taken for more ‘award’-based grant applications. I personally like the slightly longer format (~12-15 pages) for more standard types of grant proposals as I feel like I can access more accurately what is exactly being proposed, the methodology, and how the investigator thinks about the problem.
I also agree with Lawrence’s comments on evaluating the number of publications versus quality and content of those publications. It is certainly true that different specializations have different cultures around publishing.
For example, some publish with greater frequency but smaller stories, whereas others prefer to publish with lower frequency but publish larger stories. It can be hard for reviewers unfamiliar with a body of work to take the time to read all papers referenced and be able to evaluate quality and content versus frequency. So, I do like the idea to limiting submitted papers with an application to a constant number.
"In year 1 of the grant you need to have finished significant work on a 3 year grant. A little unrealistic."
My experience is both similar and dissimilar to Christine's and that described by Peter Lawrence.
The problem described for K. is one that does happen at every institution—it is very easy to be 'too successful' at first, and spend too much time spinning your wheels (writing grants, hiring people, training people) and not enough time actually writing papers. It's really about trying to balance and get enough momentum early in your career, by moving in 1st gear. In addition, realize that assistant professors are trained in their PhD to do research and write papers, and they get a job on that basis, but suddenly they have to learn how to teach courses, write grants, manage a lab, run a university, etc. It can be quite overwhelming, especially in the US system. Too bad there isn't better guidance in grad school or training when one starts a profession position on how to balance and get everything moving at once, as well as how to teach.
The idea of longer lasting proposals is very appealing. 3 year proposals are the norm in my area, and it takes 6-9 months from submitting a proposal to find it if the proposal is funded, and sometimes almost 1 year from time of submission of a grant to getting the money in place. This means that to have continuous funding, at year 2 you need to submit for a renewal, which means that you need to have significant results published at that time. Which means that in year 1 of the grant you need to have finished significant work on a 3 year grant. A little unrealistic.
I don't necessarily like the idea of giving different quality researchers a blank check—'quality' is hard to judge. But frankly, we are judged on our past performance already, and in our proposals, only the most relevant 10 papers. I like the idea of shorter proposals—I've reviewed some short proposals from the UK, and you can't really tell how they are going to do it, you really have to go on the strength of the research team and the idea. And I think you can probably judge pretty fairly from that how they will do in the future. Or at least as well as you can from a 15 page proposal, and it should be less time involved to write up. It is definitely less time to review, and I spend a significant amount of my reviewing proposals.
I work in climate change, and many prominent people in this area do not necessarily have large research programs, but rather work collaboratively with others. On the other hand, there are definitely some large prominent groups. The ideas about large groups not working efficiently is a good one—but how do we prove efficiency?
One of my biggest problems in funding is that my area of research is not straight down an NSF [National Science Foundation] program's area (a big funding source for my type of work), since I work on a new area, called biogeochemistry. The cross-cutting funding opportunities within NSF, which seek to help researchers like myself, have much lower success rates than the straight disciplinary programs.
It is also difficult to get a really new idea funded for research—it is unlikely to get good reviews if it is controversial. For that, you would need to show your idea will work first—need a first paper in any case. Reporting requirements have been increasing at the funding agencies, which is a huge burden to researchers. If we could reduce the reporting requirements that would also make the system work better. The new extra money available from the stimulus package only makes the reporting requirements worse: apparently for that money we will need to report quarterly instead of just annually.
I think it would be great if the funding agencies thought about the 'lifecycle' of a research project, and how to streamline the bureaucracy involved. This means writing time, review time, panel time and the final reporting time, and try to minimize the administrative time compared with the quality of science. Can they evaluate their reporting requirements and see which ones actually help evaluate the progress and which reporting requirements don't?
"Learning how to say 'no' to things"
Natalie and Eli -
I am not sure what start up packages in the UK are like. However, it did cross my mind in reading the article that they may be different than what we receive here in the US. For my field, we do tend to get start ups in the range that you mentioned (depending on university, research program, etc). The idea is that typically this start up will last you for ~3-4 years of running your lab, so that you can get preliminary results and papers published in order to then transition to outside funding.
In addition to thinking about the lifecycle of a research project (and grant), I think there is room for improvement in how interdisciplinary proposals are reviewed. I have had instances where a review panel that my grant application is sent to seems to have a narrow focus of disciplines very different from my own. This typically negatively affects how the grant is scored, where many of the comments reflect cultural differences between different disciplines as opposed to specific problems with the proposed research methodology.
I also agree with Natalie's comment on the type of training we receive as PhD students and postdoctoral researchers compared with what we are expected to do immediately in a faculty position. Setting up more effective training programs for researchers preparing to enter into faculty positions would be very useful.
Such programs can provide training on all the responsibilities and tasks that we don't receive training for as PhD students and postdocs such as grant writing, mentoring, teaching, accounting and resource management, but also other very important skills like time management, effective prioritization of responsibilities, and learning how to say 'no' to things.