The Scientist blog reports that a representative of the National Science Foundation (NSF) was at the annual meeting of the America Society for Cell Biology (ASCB). The NSF representative pointed out a couple of things things:
-
If your proposal describes research designed to find a cure for some disease, the NSF will not fund it. Well, duh! The NSF is about funding basic research. If you want to cure diseases, go ask the National Institutes of Health (NIH) for money. Research about human diseases and human health are not fundable by the NSF.
-
If your lab is well funded, don't expect the NSF to fund your proposal. I've heard from top-notch scientists about NSF proposals that received multiple excellent reviews, but were not funded because the lab had too much money.
These are important points. First of all, both of these appear to stem from the fact that the NSF has much less money to divvy out than the NIH. The NIH has a budget of nearly $30 billion, while the NSF has around $6 million. The NIH funds biomedical research only, while the NSF must fund many different types of scientific research -- biology, chemistry, physics, etc.
With such a diverse range of topics it must fund and much less money than the NIH with which to fund it, it makes sense that the NSF would shy away from funding research that the NIH would fund. Why "waste" their limited budget on projects that should be funded by the much richer granting agency?
Also, the NSF receives many more proposals than they can fund, and they must come up with creative ways to decide who gets funded. Many top-notch proposals will get turned down, so the NSF ensures that labs with less funding get the money that is necessary to keep their research projects above water. A well-funded lab will not be as drastically affected by an unfunded NSF proposal, but a lab that is just scratching by will suffer if their equally rated proposal does not get funded.
- Log in to post comments
Your numbers are off a bit for the NSF. If you click the link to the NSF, you can find this statement:
Something else to take into consideration is the funding rate.
NSF awards around 13% of proposals submitted. Some panels are as low as 6%. NIH funding rates are almost twice of NSF's.
A new trend is funding multi-collaborative efforts, both within the same institution, as well as between different institutions. One might say that science today is much more complex and requires in most instances a multi-disciplinary effort to answer big and even small questions. That is true.
But at the same time, there is a loss of fidelity in the process. With large multi-party proposals (I have participated in several, and currently exclusively in them) the individual sections often look somewhat shabbier when compared to what that individual would have to write for a single PI proposal. Because the whole is greater than the sum of the parts, these multi PI proposals look good and often get funded. But I am not so sure that the overall outcome is any better than individual PI grants.
NSF has some institutional programs like Engineering Research Center (ERC), and Materials Research Science and Engineering Center (MRSEC) that are multi-PI, multi-university, and multi-disciplinary. Other agencies are also following suit (DoD and I think even NIH has some centers).
Does anyone have figures on how US science funding, as percent of GDP, compares with other nations? My impression is that we are not doing too well.
@Jim:
Some western countries are nevertheless far below USA science funding... That's a very significant change when you come from oversea, and you have some real monney to do science, and you can have experiments running without repair patches...
Did a little looking around. Finland is at 3% of GDP. As far as I found out, the US is under 1% of GDP and has not been above 2% for some years.
Numbers for 2006:
EU as a whole spends about 2% on research and development.
UK 1.9%
Sweden 3.5%
and I think Finland is 3.7% in 2006, not 3%.
New Zealand 1.1%
There isn't an official estimate for India, but it works out to be around 10% -- which I am not sure makes sense.