Writing a grant with someone else's postdoc

[born_a_prof]

I am an new TT STEM prof, and before moving to this position, I had started a collaboration with a postdoc at a different university.
I am planning to write a NSF/DOD grant as a PI, in collaboration with the said postdoc.

1). Are there any special things to keep in mind in this situation, especially from strategic point of view to improve success rate? The postdoc's advisor is OK with him writing this grant.

2). I am not sure how to do budget allocation for this grant. The postdoc is receiving his salary from his advisor's funding, so not clear how to ask for his salary. Any pointers ?

General advice from anyone who has done similar things before will be greatly appreciated.

[ergative]

I've literally just done this, with two postdocs on someone else's grant. It wasn't an NSF grant, though, so I don't know how much my experience can help you. As I recall, our research office had a lot of back-and-forths with the program officers in order to make sure it was clear I wasn't budgeting for any postdoc, but for this named postdoc.

The issue of salary wasn't a problem. Since the grant required this much postdoc time, we simply asked for a pro-rated salary for that 20% time (or whatever), so they would spend 80% time on their old grant and 20% time on this one. The research office was perfectly happy calculating what that 20% salary would look like when they did our budget for us. One postdoc was quite happy about it, because it meant that the 20% salary she wouldn't be getting from her existing grant while working on mine could be used to extend her other appointment a little longer.

[Hibush]

This is one of those situations where having an excellent research office is a huge competitive advantage. They know both the program and university rules, and can figure out how to fit your goals in both sets (or devise a compliant way to get the work done).

[Kron3007]

I don't know the specifics of the grant etc., but think you should offer to cover part of the postdoc salary from the grant if it is successful.  I don't know if this is practical, but since some of the post doc's time will be devoted to the project, it only seems right to offer.

[Puget]

I do NIH grants not NSF, so take this for what it's worth, but at least for NIH people split effort between grants all the time, so it shouldn't be a problem to put 20% effort (or whatever) for the postdoc on your grant, and the rest would remain with their PI (should it be funded).

In terms of strategy, might it make sense to also include this postdoc's PI as a consultant or co-I? As a new PI, I'v been given the advice that it can help to have someone more senior on the grant.

[Liquidambar]

NSF grants are to institutions, not to individuals.  Will the postdoc's institution allow him to serve as PI on a grant?  That would be the first step.  My guess is that this won't be allowed.  (My institution doesn't allow it.)  If not, you'll have to figure out a different way to include the postdoc.  Perhaps his advisor could be co-PI, or perhaps there's a way to include him as a subcontractor.