The optimal amount of fraud is non-zero. If your financial aid process actually has 0% fraud then it is guaranteed that (1) you spent more resources vetting applicants than saving from fraud and (2) a large chunk of deserving applications were denied in the process just to be extra careful. So the 0.21% seems pretty reasonable to me.
Given the evolution of the technology, I’d say 0.21% is neither static and guaranteed to raise significantly in the to near future if the identification issues don’t get solved overnight.
If fraud detection is so expensive, maybe it's important to be more selective of who has access to the community and keep the scammers out.
"Ideal society" doesn't exist.
And how would you "be more selective" and "keep scammers out" without incurring more costs? At some point, it costs way more to prevent the last 0.1% or 0.001% of fraud than would be saved by preventing said fraud. And the cost might be paid by non-fraud being detected as false positive.
I worked in retail loss prevention, what is acceptable isn’t optimal.
Your salary was the exact kind of thing that needs to be balanced against the cost of fraud; if it was larger than the amount of fraud you prevented, then the company would have been better off just accepting the fraud as a cost of doing business. The closer you get to zero fraud, the more expensive it becomes to reduce it further (and the more likely your countermeasures will negatively impact the business in other ways), so there definitely is an "optimal" balance to be struck between fraud and preventive measures.
Simply put every living system on the planet earth has some amount of parasites. To have no parasites at all would require massive amounts of energy by the host to ensure said parasites don't exist. If the host can spend a much smaller amount of energy and ensure that 99% of parasites don't exist that is optimal thereby negating your original premise.