With the e-dollar as the unit of account, everything the central bank needs to do to have a nonzero paper currency interest rate can be done at the central bank’s cash window where banks come to deposit or withdraw paper currency from the central bank.
For good monetary policy, it is important that the central bank have control over the unit of account. And this e-dollar unit of account might have many of the aspects of a cryptocurrency—perhaps enough that it can be considered a cryptocurrency.
As far as private cryptocurrencies (like bitcoin) go, it is fine to have private cryptocurrencies perform the medium-of-exchange and store-of-value functions of money, but monetary policy requires control over the unit of account. So central banks need to retain control over the type of money that defines the unit of account—in this case the e-dollar.
Under an electronic money policy, 3 key things will insure that the e-dollar (or e-euro or e-yen or e-pound etc.) is the unit of account:
This is good.
I was about to add in Sumner's emphasis that it is the "medium of account" that must be controlled.
However, I am less and less comfortable with emphasizing the medium of account when the growth path of nominal GDP is used as a nominal anchor.
A fixed weight of gold is the medium of account--clear.
A steel ingot is the medium of account and it has no medium of exchange function -- clear
A bundle of goods is the medium of account and it has no medium of exchange and no store of value function as a bundle -- clear.
The bundle medium of account is practically the same as monetary regime that stabilizes a price index defined by the bundle -- seems right. A privatized system using index futures convertibility doesn't seem to have any alternative medium of account.
Some changing fraction of real output is the medium of account? Not so clear.
A monetary policy that stabilizes the growth path of nominal GDP is possible. A privatized system using index futures convertibility has nothing other than the unclear medium of account as some changeable fraction of real output.
And so back to Kimball, is it the unit of account that is the key to the monetary regime?