nsf-no-dup policy

Hong Ma at CSHL Delbruck mah at CSHL.ORG
Fri May 22 10:46:35 EST 1992

I support the proposal from Chris Somerville that nsf should 3 submissions a
year and shorter review time.  With my very limited experience, USDA proposals
are processed faster than NSF ones.  For 1992, the proposals submitted in 
late Jan. have now been reviwed, and some have recieved funding decisions.
The NSF proposals submitted last December are still being decided, although
I believe the panels have met.  One of the reasons might be that NSF receive
many more proposals (just a guess).  Another may be that the NSF deadline is
a recommended one; some proposals may be accepted after the recommended date.
A third reason may be the length of time outside reviewers are given to 
review the proposals.  When it comes to reviewing proposals, I believe many
if not most, people will wait until the last minute to write the reviews.
So, as long as a reasonable amount of time (for example, three weeks) is given
to a reviewer, longer time is not necessary.  People who truly cannot review
the proposals should notify NSF immediately, and another reviewer can be
contacted.  The last two of the above three possible reasons can be changed
to shorten the time of processing an NSF proposals.  I believe reducing the
reviewing time to three is possible, and I support the suggestion from June
Medford, that it should be implemented before, or at least at the same time,
as the no-dup policy.
In addition, I believe increasing the direct cost of NSF grants will go a
long way in helping the nerviousness from the community.  I think when grantees
choose NIH grant over NSF ones, it is almost always because NIH grants are 
bigger or for more years.  Chris Somerville suggested that the NSF grants should
be five years, that will certainly help.  In addition, increasing the amount
of direct cost will also help a great deal.  If NSF is to attract the best
proposal in a given field, then it should be prepared to give big awards.
One of ways this could be done, given a fixed budget, is to have a flat 
indirect cost rate, like USDA has now.  I am at an Institute which has a rela
relatively high overhead rate.  I don't know what the average is.  Something
like 30-40% would be reasonable.  For this to work, the flat rate has to be
below the average, so that the overall NSF BIO budget is spent more on
direct cost than it is now.  Of course, real increase would have to be from 
increase in the NSF BIO budget.  To that end, I am for a "unique role of NSF
in biology".  
NSF has traditionally been a strong supporter of plant basic sciences.  I hope
that will continue to be true.  To be sure, NIH has also funded some plant
projects, especially those which share certain common feature with animal
sciences.  I believe there is always going to be some overlap between NSF
and NIH, and between NSF and USDA.  As long as the majority of NSF funded
research is different from the majority of NIH or USDA research, NSF should
be able to justify its importance.  
One more comment about the no-dup policy is that it may not achieve its goals.
The current policy, as I understand it, requires that the same proposal not
be submitted to NSF and another agency at the same time.  However, it does
not require a proposal submitted to NSF NEVER be submitted to another
agency.  Therefore, the same proposal can still be submitted to other agencies
before or after it has been considered by NSF.  This could still creat an
impression of overlap or redundancy between NSF and other agencies.  

I sincerely hope both NSF and plant sciences can grow together, strengthen
each other.

Hong Ma
Cold Spring Harbor Lab.

More information about the Arab-gen mailing list