When reviewing proposals, I don't spend much time looking at the budget unless there is some particular reason to do so, such as: comments on the budget are requested/encouraged because a proposal concerns equipment or some other large-ticket item; the proposal budget total is surprisingly high or low; and/or I am curious about some aspect of the budget, e.g. how funds for a particular aspect of the research are allocated or justified. The part I care about least is what people request for summer salary. I figure that that part is between the program director and the PI's to work out, and I don't really care whether someone requests 2 weeks or 4 weeks of summer salary.
The amount of summer salary I request typically has no relationship to the actual time I spend on the research -- I always spend much more time on the project than what I can reasonably budget in terms of salary. Similarly, the random (but low) number that my university assigns as my '% effort' on a project never has any relationship to my actual 'effort', which would be nearly impossible to calculate anyway.
When constructing proposal budgets, most of my colleagues and I try to pick a 'reasonable' number that is neither too high (causing sticker shock and making one seem greedy, even if it is theoretically reasonable to request funds to make up for at least part of the 3 summer months many professors are not paid by their university) nor too low (causing people to doubt the PI's commitment to the project).
If the overall budget starts to get out of control owing to the high cost of grad/postdoc salary, fringe benefits etc., my salary request is typically the first thing to go. And even if I do keep some amount in the budget for my summer salary, if the grant is awarded and funds get tight owing to unexpected costs -- e.g. when my department mandated a raise for grad students, including for RA salaries paid from existing grants that didn't budget for this because the raise was announced without warning -- my summer salary gets whittled away because it wouldn't make sense to take the money from the amount budgeted for the actual research. It is the same for many of my colleagues as well.
It's of course nice to get paid something in the summer. I work hard in the summer, and there are various expenses involving the offspring, house, car, travel that are easier to deal with if one is paid in the summer, if only for 1 month out of the three. Or three weeks. Or two.
One of my senior colleagues refuses to tie summer salary amount to base salary. He calculates summer salary as a fixed amount that is the same no matter what the base salary of the senior personnel. That is, if he writes a proposal with a junior colleague, they both get the same summer salary. His philosophy is that they both work as hard, so why should he get paid more, at least in terms of grant-generated funds? From what I've seen, this approach is rather rare, and most people prorate summer salary requests to their 9-month base salary.
As my salary has increased over the years, I find that I ask for less summer salary, mostly for the reasons mentioned above re. priorities in a limited budget and a wish to avoid budget sticker-shock. And some years, I don't request any summer salary, even though I don't do any less research. As long as the cost of everything keeps going up and funding agency budgets don't increase, summer salary erosion will likely continue, and some of us will be able to add "volunteer" to our list of titles.
8 years ago