The funders are paying for work that isn't being done. Grants usually specific the "percent effort" of key personnel. This not only determines the budget, but

The funders are paying for work that isn't being done. Grants usually specific the "percent effort" of key personnel. This not only determines the budget, but also factors into funding decisions. A proposal might be rejected as 'infeasible' if the PI isn't going to spend much time training the students or managing the work. These numbers usually aren't exact--people certainly spend 19 (or fewer) hours per week on things that are nominally 50% effort--but having giant off-the-books projects, like running a whole separate lab--make them completely fictional.