We find conditions for the Friedman rule to be optimal in three standa
rd monetary models. Our main contribution is to shed light on two issu
es in the literature. First, the conventional view maintains that when
money is a final good, its services should be taxed. Moreover, if mon
ey demand is interest-inelastic, its services should be taxed heavily.
We show that this view is incorrect. Second, there is an ongoing cont
roversy about whether the optimality of the Friedman rule is connected
to the intermediate goods result from public finance. We resolve this
controversy by showing a deep connection between these results.