Young Female Scientist has an interesting post up concerning Francis Collins's recent oblique comments on "soft-money" positions. One of her major points--one that she has emphasized numerous times in the past--is that it is both unfair and terrible for the scientific enterprise that post-docs are almost invariably not permitted by their institutions to serve as PIs on research project grants (i.e., R01s, R21s, R03s, P01 projects, etc).
One of her subpoints is the following:
It all seems backwards to me. Seems to me that the university should be allowed to guarantee resources IF AND ONLY IF the grant is awarded. Then the university isn't risking anything.
But NIH WON'T LET THEM DO THAT.
It's not that NIH at the administrative level won't allow universities to guarantee resources if and only if a grant is awarded. It's that study sections trash grant applications on the Investigator and Environment scoring categories when the PIs are perceived as not being genuinely independent investigators.
There are a number of reasons for this study section behavior, some good and some not so good.
Reasons for this behavior include preventing lab heads with large labs from further increasing the size of their empires through R01s being applied for by "junior PIs" who are not independent, and are continuing to work within the research program of the large-lab head. Another reason is that study sections perceive that this forces to institutions to make the decision whether to devote resources to a new PI with the understanding that they have proven their abilities in the past via substantial publication records and have excellent plans for future independent research, rather than exist in some "exploited" soft-money role.
Another reason is that essentially all members of study sections are tenure-track or (now more frequently) tenured faculty members (or the equivalent) and they want to protect their privileged status as such, by limiting RPG awardees to other tenure-track (or equivalent) investigators, and not allowing perceived rabble in the door.