I wouldn’t be surprised if the “grant cancellations” alone cost dramatically more (measured as NPV today of lost future tax revenue) than the entire reported DOGE savings.
Here's a report[1] arguing that $1 spent on an NIH grant generates $2.56 of economic activity---and these grants largely aren't even meant to have short-term payouts. Block grants for (e.g.,) preventative care or education probably have huge returns too: the ER and prison are expensive!