14 Comments
2011 Success Rates, Applications, and Investigators
Well, the numbers for 2011 (fiscal year) are in. Here are a few facts about the applications and people we supported in 2011.
| Overall success rates for research project grants fell compared to 2010. | 18% |
| Success rates for new investigators were equal to established investigators submitting new applications. | 15% |
| The representation of women NIH investigators remained the same as in 2010. | 29% |
| Women’s success rates were equal to men for new applications. | 15% |
| Our commitment to supporting the individual investigator remains strong, with R01s and R37s representing a significant percentage of all research grants. | 60% |
| The average size of R01-equivalent grants increased slightly compared to 2010. | $408,594 |
| The average size of a center grant fell by 6% compared to 2010. | $1,863,037 |
| Number of institutional training grant applications continued to decline, from a peak in 2005. | 686 |
You can find all of these data and more in the NIH Data Book, which was recently updated with 2011 data. The Data Book is the first place to check when you are looking for summary statistics on NIH awards. The data and charts are exportable, making for easier reports, presentations, or blog posts.
There are a few things I wanted to note about the success rate data. A number of factors contributed to the lower RPG success rates in 2011. One of the most obvious was an 8% increase in the number of competing RPG applications. We received a record 49,592 applications. It was a busy year for everyone in the extramural community!
The slow, but steady, increase in the representation of women among NIH-supported investigators continues, and their success in applying for new grants has been the same as men’s (albeit slightly lower for renewal awards). Despite these increases, their representation in the NIH-funded pool of investigators still has not reached their level of representation among U.S. citizen postdoctorates in the biomedical sciences (approaching 50%). See my April 27, 2011 blog for more of my thoughts on this topic.
Over the next few months, I’ll post more information about these data and others in the Data Book. In the meantime, take a look.
Grant success rate has been a topic of interest for me as a researcher for 3 decades. Success rates used to be between 70-90%. It has been a depressing picture of decline over most of that period and NIH has NEVER responded to my comments and queries about the decline and their failure to understand the critical issues which would help understand their statistics. We need to know how many individual Principal Investigators are funded, or how many proposals individual investigators submit.
Expressing success rate as a fraction of submitted applications completely fails to define the success rate of individual applicants. There doesn’t seem to be a statistic indicating how many individual researchers there are relative to the number of funded grants. It isn’t clear if this ratio has remained constant, indicating a significant decline in the efficiency of the funding process, or if the funding rate has declined because there are significantly more researchers applying for funding. Most researchers today will agree that they have to write more grant proposals and spend more time writing grant proposals than they had to in the past. Many now recognize that a 20% success rate means that researchers need to write 5 proposals to get one funded. This means that the opportunities (time) for conducting research has declined, even for top-rated investigators.
It is easy to model the relationship between the number of investigators, the number of awards, the number of applications and the percentile cutoff. It can be easily shown that changing the percentile cutoff increases the number of applications per individual even if the number of investigators and the number of awards remains constant. Could declining percentiles be contributing to the downward spiral of grant success rate? And thus the downward spiral of American Science?
The information doesn’t seem available to make this judgement. It should be made available. As things currently stand, there seems to be no effort to look at the problem of diminishing scientific efficiency.
What impact will having dropped the R01 pay line at NHBLI from 16% to 10% have? thats a 37.5% reduction. How does this fit with NIH receiving a 0.8% increase in funding by congress?
I agree entirely with John’s comments. The ‘increase’ in applications that is being ascribed as one of the reasons for the drop in the success rate of R01’s may very likely represent a larger number of applications being sent by the same number of people. I personally sent 14 grants out in the past 3 years and got 3 funded, which amount to about an R01 and a half. Like many, I am sending multiple applications and trying to secure any dollar I can in order to keep my lab going. This is probably a significant source for the increasing numbers of applications – desperation to get funded – and not necessarily increased scientific activity as the NIH authorities seem to portray. Bottom line, we need better statistics if we are to understand and address this problem.
I agree with the need for data on funding per initial submission, or per submission overall. For example, this would be very helpful for the yearly talk I give to postdocs at my institution about money, where it is always challenging to present success rates. It’s even more complicated now as success of new applications seems to have risen slightly, but this may just reflect the loss of A2’s from the application pool.
I expected the women representation as NIH investigators to rise this year, glad to know at least 30% are represented by women now
I agree with the need for data. nodding
Real technicalities lie in maintaining the success rate which is currently declining to keep another aspect to limits. Datapoint discussed seems to be agreeable.