If Tim Birkhead is worried about the waste involved in failed grant proposals now, he had better hope that the funding councils don't include research grant income in a metrics-based system.
One reason why 70 per cent of applications fail is because people believe the benefits of getting a grant are big enough, so it is worth submitting proposals that have only a 30 per cent chance of winning funding. Under a metrics system, the benefits of getting a grant go up, since you get both the direct payment and a rise in your future funding council income. The stronger incentive to get a grant will surely lead to more applications and an even higher failure rate.
Richard Green
Birmingham University