> Spend the budget or next time people will ask why you need all that money when you didn't spend it last time.

I've always heard of this nugget of wisdom but never really understood it. By punishing those who underspend (by making the next application harder), wouldn't you incentivise inflated research costs, or worse, fraud. Seems like a quick path to a positive feedback loop towards the degradation of trust in academic spending, leading to "poor government efficiency".